General Transparent Solubility Prediction using Abraham Descriptors

Making solubility estimations for most organic compounds in a wide range of solvents freely available has always been a main long term objective for the Open Notebook Science Solubility Challenge. With current expertise and technology, it should be as easy to obtain a solubility estimate as it is now to get driving directions off the web.

Obviously this won't be attained purely by exhaustive measurements, although we have been focused on strategic measurements over the past two years. In parallel, we have been constantly evaluating the various solubility models out there for suitability.

Although there are several solubility models available for non-aqueous solvents, our additional requirement for transparent model building has proved surprisingly difficult to satisfy.

From this search, the Abraham solubility model [Abraham2009] floated to the top, with an important factor being that Abraham has made available extensive compilations of descriptors for solutes and solvents. In addition the algorithms used to convert solubility measurements to Abraham descriptors (a minimum of 5 different solvents per solute) has allowed us to generate our own Abraham descriptors automatically simply by recording new measurements into our SolSum Google Spreadsheet. These can be obtained in real time as well.

This approach permitted us to provide predictions for a limited number of solutes in a wide range of solvents and we have included these predictions in the past two editions (2nd and 3rd) of the ONS Challenge Solubility Book.

Coming at the problem from a different approach, Andrew Lang has also been trying to predict solubility using only open molecular descriptors, mainly relying on the CDK. Since our most commonly used solvent has been methanol, Andy recently generated a web service to predict solubility in that solvent.

By combining these two approaches, Andy has now created a modeling system that can not only generally predict solubility in a wide range (70+) of solvents - but it can also provide related data that can be used for modeling other phenomena such as intestinal absorption of a drug or crossing the blood-brain barrier.[Stovall 2007]

The idea is to use a Random Forest approach to select freely available descriptors to predict the Abraham descriptors for any solute. A separate service then generates predicted solubilities for a wide range of solvents based on these Abraham descriptors. I'm using the term "freely available" because - although the CDK descriptors and VCCLab services are open - the model requires 2 descriptors only available from ChemSpider (ultimately from ACD/Labs).

Here is an example with benzoic acid. As long as the common name resolves to a single entry on ChemSpider, it is enough to enter it and it automatically populates the rest of the fields, which are then used by the service to generate the Abraham descriptors.

Hitting the prediction link above will automatically populate the second service and generate predicted solubilities for over 70 solvents.

This approach of allowing people to access these components separately can be useful. It can be instructive to manually play with the Abraham descriptors directly to see how predicted solubilities are affected. There are also situations where one has experimentally determined Abraham descriptors for a solute and bypassing the descriptor prediction step is required.

However, for those who prefer to cut to the chase, a convenient web service is available where the common name (or SMILES) of the solute is entered and the list of available solvents appears as a drop down menu.

Now here is where I think the real payoff comes for accelerating science with openness. Andy has also created a web service that returns the predicted solubility in molar as a number from common names (or SMILES) for solute and solvent via the URL. For example click this for benzoic acid in methanol. The advantage here is that solubility prediction can be easily integrated as a web service call from intuitive interfaces such as a Google Spreadsheet to enable even non-programmers to make use of the data. Notice that the web service provided in the fourth column for the average of measured solubility values enables an easy way to explore the accuracy of specific predictions.

Such web services could also be integrated with data from ChemSpider or custom systems. If those who use these services feed back their processed data to the open web, it could take us a step closer to automated reaction design. For example consider the custom application to select solvents for the Ugi reaction. Model builders could also use the web services for predicted and measured solubility directly.

A while back we explored using Taverna for MyExperiment to create virtual libraries of SMILES. Unfortunately we ran into issues with getting the applications developed on Macs to run on our PCs. This might be worth revisiting as a means of filtering virtual libraries through different thresholds of predicted solubility.

Andy has described his model in detail in a fully transparent way - the model itself, how it was generated and the entire dataset can be found here. We would welcome improvements of the model as well as completely new models based on our dataset using only freely available tools.

It should be noted that when I use term "general" it refers to the ability for the model to generate a number for most compounds listed in ChemSpider. Obviously compounds that most closely resemble the training set are more likely to generate better estimates. Because of our synthetic objectives using the Ugi reaction we have mainly focused on collecting solubility data for carboxylic acids, aldehydes and amides either from new measurements or from the literature.

Another important point concerns the main intended application of the model: organic synthesis. Generally the range of interest for such applications is about 0.01 - 3M. This might be very different for other applications - such as the aqueous solubility of a drug, where distinctions between much lower solubilities may be important.

For a typical organic synthesis, a solubility of 0.001M or 0.005M will probably translate as effectively insoluble. This might be a desired property for a product intended to be isolated by filtration. On the other end of the scale knowing that a solubility is either 4M or 6M will not usually have an impact on reaction design. It is enough to know that a reactant will have good solubility in a particular solvent.

Given the above considerations for intended applications and the likelihood that the current model is far from optimized, the predictions should be used cautiously. We suggest that the model is best used as a "flagging device". For example, if a reaction is to be carried out at 0.5M, one may place a threshold at 0.4M for the predicted values of reactants during solvent selection, with the recognition that a predicted 0.4M may be an actual 0.55M. A similar threshold approach can be used for the product, where in this case the lowest solubility is desired. A practical example of this is the shortlisting of solvents candidates for the Ugi reaction.

Another example of flagging involves identifying the outliers in the model. These can be inspected for experimental errors and possibly remeasured. Alternatively outliers may shed light on the limitations of the model. For example we have found that the solubility of solutes with melting points near room temperature can be greatly underestimated by the current model. This may be an opportunity to develop other models which incorporate melting point or enthalpy of fusion.[Rohani 2008]

Although it is possible that better models and more data will improve the accuracy of the predictions, this can be true only if the training set is accurate enough. Based on conversations I've had with researchers who deal with solubility, reading modeling papers and our own experience with the ONS Challenge I am starting to suspect that much of the available data just isn't accurate enough for high precision modeling. Models using data from the literature are especially vulnerable I think. Take a look at this unsettling comparison between new measurements and literature values (not to mention the model) for common compounds.[Loftsson 2006] Here is a subset:

I have also made the point in detail for the aqueous solubility of EGCG. Could this be the reason that so many different solubility models using different physical chemistry principles have evolved and continue to co-exist?

The situation reminds me a lot of the discussions taking place in the molecular docking community.[Bissantz 2010] The differences in calculated binding energies are often small in comparison with the uncertainties involved. But docking can still be used as one tool among others to find drug candidates by flagging a collection of compounds above a certain threshold binding energy.

Resveratrol Thesis on Reaction Attempts

A few days ago Andrew Lang suggested to Dustin Sprouse that he submit his thesis to the Reaction Attempts database. Like many undergraduates Dustin put in a lot of time and effort in doing experiments and writing up his results but didn't have quite enough time to obtain all that would have been required for a traditional publication.

A thesis is an unusual document within the context of scientific communication. Unlike a peer reviewed paper, it may contain a large number of "failed experiments" and a substantial amount of speculation. Although it is not quite as detailed as lab notebook, there is often plenty of raw data and details about how failed or ambiguous experiments proceeded.

In Dustin's case we felt that there was enough information provided to include his thesis in Reaction Attempts. In addition, his thesis was accepted by Nature Precedings, thus providing a convenient means of citation.

The first component of the Reaction Attempts project is to quickly abstract the most basic information from synthetic organic chemistry reactions. This includes the ChemSpiderIDs and SMILES from the reactants and target products and brief notes about conditions and outcomes. We are especially interested in failed or ambiguous experiments because these have almost no chance of being communicated and indexed in the traditional systems. When attempting to carry out a reaction, it can be just as useful to know what doesn't work - and more specifically how it doesn't work.

The second component of the project is dissemination. Because the information is encoded semantically, it can be automatically converted to both human and machine readable formats.

One human interface consists of a PDF book (also as a hard copy), with the option of selected reactions specified by listing CSIDs of reactants in the URL. For example Dustin's reactions can be presented selectively here. We also have a Reaction Explorer, where reactants or products can be selected from a dropdown menu or via a substructure search.

We also provide live XML feeds so that others can create applications easily from machine readable data. For example one could create reaction chains automatically, which will occur whenever we enter reactions from multi-step syntheses like Dustin's - based on the synthesis of resveratrol.

I know that Peter Murray-Rust has been very active in automatically abstracting information from chemistry theses. It would be interesting to see how that approach would work for this thesis, especially with the failed experiments. Reducing a page or two of text into only the most salient bits of information manually required a level of judgement that I imagine would be tricky to do automatically.

Secrecy in Astronomy and the Open Science Ratchet

Probably because of the visibility of the GalaxyZoo project, I think several of my colleagues and I have been under the impression that astronomy is a somewhat more open field than chemistry or molecular biology. It was easy to rationalize such a position because patents are not an issue, as they clearly are in fields which rely more on invention than discovery. However, after reading "The Case for Pluto" by Alan Boyle, I am left with a much different impression.

This book does an excellent job of covering the recent debate over Pluto's designation as a true planet. A key trigger for this debate has been the discovery of dwarf planets with sizes very close to that of Pluto. However, these discoveries did not occur without controversy.

The story of the controversy regarding the discovery of Haumea is a particularly good example (starts on p. 108 of the book - a good summary also on Wikipedia). Starting in December 2004 Michael Brown at Caltech discovered a series of new dwarf planets. Instead of immediately reporting his team's discoveries, he worked in secrecy until July 20, 2005 when he posted an online abstract indicating the discoveries would be announced at a conference that September. However, on July 27, 2005 a Spanish team led by José Luis Ortiz Moreno filed a claim with the Minor Planet Center for priority in discovering one of these dwarf planets. This forced Brown's hand in disclosing his team's other discoveries within days - much sooner than he had anticipated.

Apparently this stirred up a great controversy in the community and officially no name was associated with the discovery, although the Spanish team's telescope at Sierra Nevada Observatory was recognized as the location of the discovery. However, Brown was allowed to select the name Haumea for the dwarf planet.

Even though the Minor Planet Center accepted Moreno's submission, most reports seem to side with Brown. The main argument is no less than academic fraud on Moreno's part because he accessed public telescope logs and found some of Brown's data. It was as simple as Googling the identifier that Brown inserted in his public abstract.

If Moreno had hacked into a private computer from Brown's team I can understand fraud. But is it fraud to access public databases? We chemists do that all the time - reading abstracts from upcoming conferences to try to glean what our competitors are up to. That hasn't stopped anyone from submitting a paper or patent.

Secrecy only works if everyone competing follows the same rules. If there is a rule that planet discoveries must be made at conferences or by formal publication then this could not have happened. Moreno's submission to the Minor Planet Center should have been rejected if such a rule existed. If there is a rule that telescope logs should not be accessed then why make them public and indexed on Google?

Now there may exist field specific conventions. I don't know what they are in the case of discoveries such as these but here is an interesting quote from Michael Brown's Wikipedia page:

When asked about this online activity, Ortiz responded with an email to Brown that suggested Brown was at fault for "hiding objects," and said that "the only reason why we are now exchanging e-mail is because you did not report your object."[3] Brown says that this statement by Ortiz contradicts the accepted scientific practice of analyzing one's research until one is satisfied that it is accurate, then submitting it to peer review prior to any public announcement. However, the MPC only needs precise enough orbit determination on the object in order to provide discovery credit, and Ortiz et al. not only provided the orbit, but "precovery" images of the body in 1957 plates.

It seems to me that there is a clash of what are the conventions in the field. Certainly the Minor Planet Center did not recognize the convention of peer review before public disclosure. They only required sufficient proof for the discovery.

One way to look at this story is that Moreno acted more openly than Brown by disclosing information before peer review. This action forced Brown to disclose scientific results much more quickly than he had anticipated.

In a sense this is a type of Open Science Ratchet. The actions of scientists that are most open set the pace for everyone else working on that particular project, regardless of their views on how secretive science should be.

Imagine how the scenario would have played out if one of the groups had used an Open Notebook. On December 28, 2004 everyone with a stake in the search for planets would have had the opportunity to know that a very significant find had been made. There were still details to work out - and the Brown group might not be the first to do all the calculations to completely characterize the discovery. Certainly it would affect what other researchers did - even if they were completely opposed to the concept of Open Science.

Essentially secrecy in this context is an all-or-nothing gamble. Everyone is free to not disclose their work until after peer reviewed publication. In some cases the discoverer will get full credit for the discovery and the complete analysis. But in other cases another group working in parallel will publish first and leave nothing to claim.

As scientists become more open, it is likely that their ability to claim sole priority for all aspects of a discovery will be reduced. However, they will retain priority for the observations and calculations that they made first.

The more open the science, the faster it happens. And because of the Open Science Ratchet, a few Open Scientists scattered across various fields could have a larger hand than expected in speeding up science.

Methanol Solubility Prediction Model 4 for Ugi reactions in the literature

Since non-aqueous solubility measurements have not become part of the standard characterization of organic compounds, it is not surprising that all the data we have for Ugi products originate from measurements that we made on our own compounds.

Since methanol is our most common solvent, Andrew Lang has collected the measurements that we have with values from the literature for a range of compounds, including our Ugi products, to generate a web service returning a predicted solubility based on a submitted SMILES string. The model (Model 4) was derived from a Random Forest algorithm, using molecular descriptors supplied by the CDK and VCC.

It would be nice to be able to test the model's ability to predict what will happen if a Ugi reaction is carried out in methanol. Although the actual solubility of Ugi products in the literature is typically not reported, reading the experimental sections in papers can still provide some validation of the model in some cases.

For example, consider the following Ugi products synthesized recently by Lezinska (Tetrahedron 2010)


Note that these images represent the azide group not following the octet rule. It is necessary to represent the structure SMILES without charges because the CDK and VCC web services used by the model do not process charges correctly. Stereochemistry also cannot be used and this can be removed from the SMILES simply by deleting slashes. Thus for the two molecules above the SMILES to be submitted to the prediction web service are:

O=C(NC1CCCCC1)C(Cc2ccc(C)cc2)N(c4ccccc4C(=O)c3ccccc3)C(=O)C(Cc5ccccc5)N=N#N
AND
O=C(NC1CCCCC1)C(C(=O)c2ccccc2)N(Cc3ccc(C)cc3)C(=O)C(C)CCN=N#N

The predicted methanol solubilities are respectively 0.004 M and 0.03 M.

Now if we look at the details in the experimental section, both of these Ugi products were synthesized in methanol at a limiting reactant concentration of about 0.1 M. Even though this is much more dilute than the usual 0.5-2.0 M generally recommended for Ugi reactions (Domling 2000), the products still precipitate and can be filtered off. This is consistent with the predicted solubilities above and the model would have suggested ahead of time that methanol might be a good solvent for isolation of the products by precipitation.

So far these are just anecdotal results but it does illustrate that solubility models can be evaluated without explicit determination of solubility in the literature.

Reaction Attempts Explorer

Two months ago I reported on the Reaction Attempts project and the availability of the summary as a physical or electronic (PDF) book. The basic idea behind the project is to collect organic chemistry reaction attempts reported in Open Notebooks. This would include not only successful experiments but also those which could be categorized as failed, ambiguous, in progress, etc.

The book was organized with reactants listed alphabetically. In this way one could browse through summaries of the types of reactions being attempted by different researchers on a reactant of interest. There might be information there (what to do or what to avoid) of some use for a planned reaction. At the very least one could contact the researcher to initiate a discussion about work that had not yet been published in the traditional system.

Andrew Lang has just created a web-based tool to explore the Reaction Attempts database in much more sophisticated ways.

Here are some scenarios of how one could use it. On the left hand side of the page is a dropdown menu containing an alphabetically sorted list of all the reactants and products in the database. Lets select furfurylamine.


This immediately informs us that there are 230 reactions involving furfurylamine and it lists the schemes for all these reactions upon scrolling down. That's still a bit hard to process so a second dropdown menu appears populated with a list of other reactants or products involved with furfurylamine.

We now select boc-glycine and that narrows our search to 145 reactions.

Selecting benzaldehyde from the third dropdown menu narrows the search further to 61 reactions.

The final dropdown menu contains a short list of only isocyanides and thus all represent attempted Ugi reactions. Selecting t-butyl isocyanide gives us 56 reactions.

That means that these same 4 components were reacted together 56 times. Looking at the various reaction summaries will show that some of these are duplicates for reproducibility and others vary concentration and solvent and the effect on yield is included. This particular reaction was in fact the subject of a paper on the optimization of a Ugi reaction using an automated liquid handler.

Now here is where the design of the Explorer comes in handy. We might want to ask if the reaction proceeds as well with the other isocyanides. All we have to do is switch the final dropdown menu to ask what happens when we go from t-butyl to n-butyl isonitrile. There is a single attempt of this reaction and it is "failed" in the sense that no precipitate was obtained from the reaction mixture. This doesn't mean that the reaction didn't take place - it might be that the Ugi product was too soluble. We can quickly inspect that the concentration and solvent are in line with conditions that allowed precipitation of the t-butyl derivative.

OK lets see what happens with n-pentyl isocyanide.

It looks like it behaves just like n-butyl isocyanide: another single non-precipitation event. What about benzyl isocyanide?

This time we do get the Ugi product from a single attempt. Note the lower yield compared to the t-butyl isocyanide under similar conditions.

What about with cyclohexyl isocyanide?

This time we hit an experiment in progress. A precipitate was obtained but it was not characterized. We can click on the link to the lab notebook page (EXP232) to learn more about how long it took for the precipitate to appear but there are not enough data to draw a definite conclusion about the successs of the reaction. However, based on the results from the other precipitates in this series it is probably encouraging enough to repeat and characterize the product.

There are other sources of information here. Clicking on the image of the Ugi product takes us to its ChemSpider entry. In this case the only associated data relates to this reaction attempt.

Lets look at another scenario: reactions involving aminoacetaldehyde dimethyl acetal.

In this case we find the intersection of two Open Notebooks. The first reaction comes from Michael Wolfle from the Todd group.

The second comes from Khalid Mirza from the Bradley group.

In order to learn more about the nature of the overlap we can use the substructure search capabilities of the Reaction Explorer. Simply click on the image of the acetal and the ChemSpider entry pops up. Now click on the copy button next to the SMILES for the compound.

Paste the SMILES into the SMARTS box of the Reaction Explorer.

We get 13 reaction attempts for this query - the two we found earlier and the rest corresponding to attempts by Michael Wolfle to synthesize praziquanamine.

We learn that one connection between these two notebooks involves different attempts at synthesizing praziquantel.

Hopefully this demonstrates the value of abstracting organic chemistry reaction attempts from Open Notebooks into a machine readable format. Contributions to the database require only the ChemSpider IDs of the reactants and product and a link to the relevant lab notebook page. Reaction schemes are automatically generated by the system. More on the Reaction Attempts project here.

IGERT NSF panel on Digital Science

On May 24, 2010 I was part of a panel in Washington for the NSF IGERT annual meeting. As I mentioned previously, it is encouraging to find that funding agencies are paying more attention to the role of new forms of scholarship and dissemination of scientific information.

My co-panelists included Janet Stemwedel, who talked about the role of blogging in an academic career, Moshe Pritzker, who made a case for using video to communicate protocols in life sciences and Chris Impey, who demonstrated applications of clickers and Second Life in the classroom.

We only had 10 minutes each to speak so the presentations were basically highlights of what is possible. Still, it was enough to stimulate a vigorous discussion with the audience. There was a bit of controversy about the examples I used to demonstrate the limitations of peer review in chemistry. People can misinterpret what we are trying to do with ONS - it certainly doesn't include bringing down the peer review system (not that we could anyway). But we have to face the situation that peer review does not validate all the data and statements in a paper. It operates at a much higher level of abstraction. Providing transparency to the raw data should work in a synergistic way with the existing system.

My favorite part of the conference was easily Seth Shulman's talk on the "Telephone Gambit". Ever since reading his book, I have been using the story of how carefully reading Bell's lab notebook has forced us to revise the generally accepted notion of how the telephone was invented. Seth's presentation was truly captivating because he explained not only what was done but also what motives were at work to deceive and obfuscate. This cautionary tale is still very much relevant to science and invention today - and highlights how transparency can mitigate against this type of outcome.

IGERT Open Notebook Science

View more presentations from Jean-Claude Bradley.

Use of ONS to protect Open Research: the case of the Ugi approach to Praziquantel

As we were collecting reactions from The Synaptic Leap for the Reaction Attempts project, Andrew Lang noticed that there might be a quick synthetic route to praziquantel via a Ugi reaction. I researched it further and found a paper (Kim et al 1998) where Ugi product 1 was indeed converted to racemic praziquantel via the Pictet-Spegler cyclization.


Using Beilstein Crossfire the only synthesis of 1 I found involves a multi-step amidation strategy. But this compound should be accessible in one step from commercially available starting materials via a Ugi reaction (shown above). Since all the starting materials are liquids we have some flexibility with solvent choice. Khalid first tried it in methanol EXP258 a few weeks ago but did not get a precipitate. He was going to monitor it by NMR next to see if the problem was high solubility of the Ugi product or with the reaction itself.

It was therefore with great interest that I read Mat Todd's report this morning on The Synaptic Leap that a German patent had been issued on this Ugi strategy to praziquantel. (TSL didn't provide a means of leaving a comment so I edited the page - which made me the author of that post but actually Mat wrote it)

I have often mentioned during my talks that Open Notebook Science could be used not only in a defensive manner to claim academic priority - but also as an offensive tactic to block patent applications. A company attempting to prevent the commercial exploitation of rival inventions has a few options. Where applicable, it can buy up an existing patent pool with the intention of sitting on it. For new inventions, it can do research and try to file patents before their competitors. But this is a costly process and it may make more sense to simply publish the inventions to create disclosed prior art, thereby blocking patent applications of their competitors.

But - as I and many others have discussed - the current publication system is not optimally suited for the purpose of simply disclosing and communicating science. Not only is it generally slow but the traditional article format requires a narrative of some sort - rarely can single experiments be published. This means that much (if not most) of research done by an individual or group will never be disclosed.

For these reasons I think that keeping an easily discoverable Open Notebook for projects designed to block patent submission by competitors makes a lot of sense - both economically and from a workflow perspective. Since researchers already have to keep a lab notebook, making it public doesn't impose the added time that writing an article or patent will require.

In this specific example of praziquantel we were too late. But if we had recorded this experiment a few years ago it might have worked to block Domling's patent. Now, it isn't clear to me that EXP258 would have been enough to do that. The strategy to make praziquantel via a Ugi reaction was clearly stated but the experiment was not conclusive. However, since Domling reported that methanol worked I am sure that we would have had the "reduced to practice" evidence in the notebook shortly.

Above I used a company as an example of a party motivated to disclose inventions to protect their interests. In our case it would not be a company but rather the entire Open Science community. It is in our best interest to keep our scientific territory as unencumbered by patents as possible. Keeping Open Notebooks might be one of the simplest means of ensuring that.

Consider a humanitarian organization that might want to manufacture praziquantel. I haven't researched it but presumably the Domling patent was filed in a number of countries beside Germany. In order to consider using the Ugi strategy, the organization would now have to deal with the patent holder. This might be the factor that makes this route untenable. Patents have proven to be problematic for humanitarian aid - even in the simple case of providing food.

But all is not lost. In addition to offering a simple 2-step synthesis of praziqantel, the Ugi route offers an easy way to make large libraries of analogs. Optimally we would like to work with someone who has experience with docking praziquantel. It might be interesting to screen not only the praziquantel analogs but also the uncyclized Ugi products themselves. When we did this for malarial enoyl reductase inhibitors (D-EXP005) we found that we did not need to cyclize to obtain compounds predicted to bind. This ultimately led to active compounds.

OpenSciNY Open Notebook Science Talk

On May 14, 2010 I presented on Open Notebook Science at the OpenSciNY conference at the New York University Bobst Library. I introduced the topic by telling a few stories about how new forms of communication are affecting how we think about concepts like "scientific precedent", "peer review", "scientific publishing" and "scientific scholarship". At the end I spoke about archiving Open Notebook Science projects and showed the physical copies of both the Reaction Attempts and ONS Solubility Challenge books.

Margaret Smith did a wonderful job of organizing the conference with a very interesting line-up of speakers: Heather Joseph, Antony Williams, Elizabeth Brown and David Hogg. We formed break-out sessions at the end to discuss with the attendees concepts around Open Science. I was part of the session on Promoting Open Science.

The tone at this and other similar conferences I have attended recently is probably best described as cautiously optimistic and focused on what is possible. The Open Science movement - at least as far as it is reflected by the people I know - does not seem to be driven by zealots or idealists trying to get everyone to drink the cool-aid. It is just a bunch of people who see opportunities to do things in better ways as new tools become available - and they can't find a credible reason not to do them.

Check here on FriendFeed for updates about links to recordings, slides, etc.

My presentation below:

OpenSciNY Open Notebook Science

View more presentations from Jean-Claude Bradley.

The Scientist Article on Electronic Lab Notebooks

Amber Dance has written an article in The Scientist (2010-05-01) Digital Upgrade: How to choose your lab’s next electronic lab notebook. This is basically a quick overview of different Electronic Lab Notebooks (ELNs) that should be helpful for people researching what is currently available in that space.

There was some coverage of Open Notebook Science and Steve Koch and I were quoted. Ironically my contribution appeared in the "Cons" section :)

Pros

• The format is unconstrained—you can set up any categories, and as many users and pages, as you want—and fast to set up.
• Open notebooking attracts collaborators. Koch counts three collaborations that wouldn’t have happened if he weren’t on OpenWetWare. And his students build professional networks well before they author a paper.

Cons

• Wikis were not designed with scientific data in mind. For example, it’s hard to make a table, Koch says.
• Open notebook science “does limit where you can send your work,” says Jean-Claude Bradley, a chemist at Drexel University in Philadelphia, who also uses an open wiki notebook. His lab sticks to journals that accept preprints.
• Posting online voids international patent rights, although US patents are still possible.

In my opinion, one of the biggest "Pros" wasn't listed in that section: the free cost. (That was mentioned elsewhere though) When you see the costs of some of these other commercial systems, that has to be a factor for many people trying to make a decision.

If privacy is an issue wikis can certainly be made private, although I'm not sure if that is possible on OpenWetWare. It can be done for $5/month on Wikispaces, the wiki we use for lab notebooks - although then it wouldn't be Open Notebook Science.

Concerning Steve's Con of wikis being difficult to use to store data, that is true. However combining the use of a wiki with Google Spreadsheets has completely resolved that issue for us. With our ability to automatically export an archive of the notebook (as HTML) and spreadsheets (as XLS) into an integrated archive, the two platforms operate essentially as if they were a single system.

Visualizing Chemistry in Second Life ACS Talk Recording

The American Chemical Society has processed the recording of the talk that Andrew Lang and I gave at the Spring 2010 ACS meeting in San Francisco on March 23, 2010:
Visualizing Chemistry in Second Life

ChemSpider SyntheticPages

I recently mentioned the Reaction Attempts project, which aims to collect organic chemistry experiments - especially those that are "failed", in progress or somehow incomplete.

For reactions where the desired product has been obtained and fully characterized, ChemSpider SyntheticPages also offers a very convenient publication vehicle. As I mentioned previously there is a need for enabling the publication of single experiments, especially when these are unlikely to become part of a traditional article.

We are in the process of submitting suitable reactions from the UsefulChem project to CS|SP. This will require some re-formatting of procedures and characterization data as they currently appear in the lab notebook.

Here is an example of one of our Ugi reactions: SyntheticPage 406 (UCEXP176C)


A nice feature of these pages is the automatic rendering of 2D structures upon hovering on top of chemical names.


Here are a few more reasons to use ChemSpider SyntheticPages:

* ChemSpider SyntheticPages takes you directly to a procedure. When you get a hit - you get a procedure.
* ChemSpider SyntheticPages provides information that may not generally be found elsewhere, such as frequently encountered problems, trouble-shooting tips, the number of times the reaction has been carried out, scale-variation etc.
* ChemSpider SyntheticPages is the only interactive chemistry database. Information is constantly updated and validated by comments from the user community (Peer Review in the Public Domain™).
* ChemSpider SyntheticPages can provide you with the most up-to-date method, we aim for 95% of submissions to be processed within 48 hours of submission.
* ChemSpider SyntheticPages is free of charge.

[Disclaimer: I am a member of the editorial group at CS|SP]

The Synaptic Leap Experiments on Reaction Attempts

Andrew Lang and I recently reported on the first edition of the Reaction Attempts book and database. Part of the motivation for this was to structure the experiments from the UsefulChem project in both a machine readable format and one that could be browsed as a physical copy. However, we also had in mind the easy integration of other open experiments, especially those labeled as "failed", since these are unlikely to be found by searching conventional reaction archives.

As a demonstration, we have added a series of experiments from The Synaptic Leap, which Michael Wolfle (working as a post-doc with Mat Todd) has posted. All of these reactions involve intermediates in the synthesis of praziquantel, which is a major focus of the Todd group. One group of these reactions involved the attempted synthesis of praziquanamine via a Pictet-Spengler cyclization. Most of these are failed attempts and one successful one.

Adding these experiments to Reaction Attempts was very simple - since the minimum information required is the ChemSpiderIDs (CSIDs) of all the reactants and the product, which a hyperlink to more details. We also added a few more details provided by Michael - such as the solvent, reaction conditions and outcome.

Andy has provided a simple mechanism to pull up all Reaction Attempts for a given reactant with the following url structure:

http://showme.physics.drexel.edu/onsc/databook/ucdatabook.php?reactants=9099925

The number at the end is the CSID for the reactant. Multiple reactants can be pulled from the database by adding more CSIDs separated by commas.

Successful runs in Reaction Attempts are identified with a green check mark:


Again the main idea here is not to exhaustively abstract all pertinent information for an experiment. Rather it is to connect up researchers who are working on similar reactions. Since it requires so little effort to come up with the minimum required information we are hoping to get contributions from other sources.

We will focus next on coming up with more sophisticated ways to retrieve information - such as substructure searching or by reaction type, solvent, etc. We will also periodically publish hard copies of future Reaction Attempts editions.

NMR integration web service expanded

The ONS Challenge has extensively used a web service created by Andrew Lang to automatically calculate solubility from NMR spectra. One of the constraints of the service was that the JCAMP-DX file had to be deposited in a special folder on a server at Drexel.

Andy has now modified the script so that the JCAMP-DX file can be located anywhere on the internet. I have prepared a modified Google Spreadsheet to serve as a template for SAMS calculations (Semi-Automated Measurement of Solubility). Simply enter the url to the JCAMP-DX file in the appropriate column and fill in the ppm ranges and corresponding hydrogen numbers for the solvent and solute, and molecular weight and density data. (The predicted density of solids can be found on Chemspider). The concentration of the solute will then be automatically calculated based on an assumption of volume additivity.

The web service (which handles baseline correction) could be used for any other purpose involving the integration of spectra. Just make a copy of the Google Spreadsheet and modify.

Note that the JCAMP-DX files must be in XY format. If your instrument saves spectra in a compressed format they must be converted to XY. The desktop version of Robert Lancashire's JSpecView can be used to carry out the conversion.

This template spreadsheet also features a service in a cell to display the NMR spectrum by simply clicking on the link inside the cell. This is very handy because it obviates the need to create an HTML file which must normally accompany the JCAMP-DX file for viewing. Being able to quickly view a spectrum from a particular row within the Google Spreadsheet makes tracking data provenance very intuitive and errors easy to spot.

Reaction Attempts Book Edition 1 and UsefulChem Archive

I am pleased to report that Andrew Lang and I have published the first edition of the Reaction Attempts book. It currently contains most of the Ugi reactions from the UsefulChem project and is associated with an April 27, 2010 snapshot archive of the entire UsefulChem project, including NMR spectra, spreadsheets, images and the entire lab notebook from Wikispaces.


At 582 pages the printing cost from LuLu amounts to $26.28. Not meant to replace electronic searches, it should prove to be a handy reference book for the lab to quickly browse through what was attempted for a given reactant, what the outcome was and the researcher involved.

We are hoping to include reaction attempts from other groups in future editions. More details can be found in the preface, reproduced below:

Reaction Attempts First Edition

Data Source: the UsefulChem project

Introduction

Open Notebook Science (ONS) refers to the practice of making the full contents of a laboratory notebook and all associated raw data files available in near real time.[1] This represents an opportunity for everyone to benefit from work in progress in an open research group. However, in order to make use of the information, it must be easily discoverable. A simple strategy to increase discoverability is redundancy over multiple communication platforms.

In another project - the Open Notebook Science Solubility Challenge[2] - we published non-aqueous solubility data in the form of physical and downloadable (PDF) books.[3] Although it is possible to search the solubility database using web query interfaces, exploration of a Google Spreadsheet, an XML feed, etc.[4], having a physical copy in the laboratory has proved to be very convenient in several instances. A similar format for reactions will also be useful.

The UsefulChem Project

UsefulChem started in 2005 as an organic chemistry Open Notebook Science project with a main goal of discovering new anti-malarial agents that can be prepared by simple and cheap syntheses.[5] Most of the reactions on UsefuChem are Ugi reactions, which involve the mixing of an amine, aldehyde, carboxylic acid and isonitrile in a solvent at room temperature generally for a few hours to days.[6] The multicomponent design of the Ugi reaction and the simple reaction conditions make it ideal for exploring large virtual libraries and selecting compounds of interest to make.[7]

Isolation of the Ugi products can be immensely simpler, cheaper and readily scalable if they precipitate in pure form from the reaction mixture. To this end, much of the research in the UsefulChem project focuses on reaction conditions that lead to this outcome.[8] This is in fact the origin of the ONS Solubility Challenge discussed above.[9]

The Reaction Attempts Database

In order to look for patterns in the reaction conditions which led to Ugi product precipitation, the CombiUgiResults Google Spreadsheet was set up.[10] Reactions indexed there can be sorted by precipitation outcome, solvent, reactant, concentration, etc. and links to the laboratory notebook pages can be followed for full details. However, this sheet is designed specifically for Ugi reactions and contains columns specifically for the aldehyde, amine, carboxylic acid and isonitrile.

In order to enable the tracking of other types of reactions, the information in the CombiUgiResults sheet was reformatted into two other sheets: ReactionAttempts[11] (containing reagents and reactants) and RXIDsReactionAttempts[12] (containing reaction conditions and results, such as solvent, concentration of limiting reactant, appearance of a precipitate, yield, etc.). The two sheets are connected via the use of a common ReactionID. This format permits the representation of any type of reaction, with an unlimited number of reactants and products.[13]

By definition, any Open Notebook Science project in a work in progress. The listing of a reaction in this database only means that the researcher attempted or is in the process of attempting it. Whatever the situation, a link to the laboratory notebook page is provided, where the most recent information is available. The philosophy used here is that partial information is always better than no information at all. Thus a researcher investigating the prior use a particular reactant in a Ugi reaction might find the report that a precipitate was obtained in methanol helpful for designing their own reactions, even if the characterization of the precipitate is still pending. At the very least, knowing that a certain researcher has at least attempted a similar reaction is enough information for initiating a discussion, which may lead to valuable insights.

Reaction Attempts on Chemspider

Although SMILES[14] are provided in the spreadsheets, the primary key to identify compounds is the ChemSpider ID (CSID)[15]. This allows us to render molecule images in the book automatically. In the case of the ONS Solubility Challenge book[3], use of the CSID enables a convenient way to calculate various descriptors for displaying values in the book.

In addition, the compounds in the Reaction Attempts database are indexed on ChemSpider as two Data Sources: ReactantsAttemptedReactions and ProductsAttemptedReactions[13]. In this way a substructure search for either reactants or products will identify indexed molecules. Clicking on the Syntheses tab in the ChemSpider record for a selected molecule will then reveal a list of hyperlinks to the relevant laboratory notebook pages.

Organization of the Book

In keeping with the layout of the ONS Solubility Challenge Book, the reactants are listed in alphabetical order. Each entry displays the list of reactions where the reactant was used. This includes a scheme with all reactants and product as well as key metadata: the researcher, reaction type, solvent, limiting reactant concentration, observation of a precipitate, comments and a reference (links to the laboratory notebook page).

In this edition, only Ugi reactions are included. The reaction schemes are laid out in the following order: carboxylic acid, amine, aldehyde and isonitrile. This should allow for easy comparison between schemes within a given record. Reactions where the Ugi product was isolated and characterized are marked with a green check and the percent yield is noted. Since the Ugi products do not have simple common names, they are not included as separate entries. However, all reactions where the synthesis of a specific Ugi product was attempted can be found by looking up the entries for any of the four reactants.

Although this compilation is not exhaustive, it does cover the vast majority of reactions in the UsefulChem project to date. Future editions will include other reactions from UsefulChem and other sources.

Archive

This edition is linked to the UsefulChem data archive (ZIP)[16], (DVD)[17] and interactive hosted archive format[18], ReactionAttempts (XLS)[19] and RXIDsReactionAttempts(XLS)[20] taken on 2010-04-27.

References

1. Open Notebook Science Wikipedia Entry http://en.wikipedia.org/wiki/Open_Notebook_Science
2. Open Notebook Science Solubility Challenge Wiki http://onschallenge.wikispaces.com
3. Bradley, J.-C. First Edition of ONS Solubility Challenge Book UsefulChem Blog (2009)
http://usefulchem.blogspot.com/2009/12/first-edition-of-ons-solubility.html
4. Open Notebook Science Solubility Challenge List of Experiments page http://onschallenge.wikispaces.com/list+of+experiments
5. UsefulChem Wiki http://usefulchem.wikispaces.com
6. Ugi Reaction Wikipedia Entry http://en.wikipedia.org/wiki/Ugi_reaction
7. Dömling, A., & Ugi, I. (2000). Multicomponent Reactions with Isocyanides. Angewandte Chemie International English Edition, 39(18), 3168-3210. http://www3.interscience.wiley.com/journal/73500473/abstract.
8. UsefulChem List of Experiments http://usefulchem.wikispaces.com/All+Reactions
9. Bradley, J.-C. Open Notebook Science Challenge UsefulChem Blog (2008)
http://usefulchem.blogspot.com/2008/09/open-notebook-science-challenge.html
10. CombiUgiResults Google Spreadsheet http://spreadsheets.google.com/ccc?key=plwwufp30hfpUERhse9y5Kw
11. ReactionAttempts Google Spreadsheet
http://spreadsheets.google.com/ccc?key=0Ak1R8T6wt4YQdG9NejNLcDNUMkVBVURGM01TR0NxdXc
12. RXIDsReactionAttempts Google Spreadsheet
http://spreadsheets.google.com/ccc?key=0Ak1R8T6wt4YQdGVENVFMWjdzaGd2REJTTnA4RG5vblE
13. Bradley, J.-C. Reaction Attempts on ChemSpider UsefulChem Blog (2010)
http://usefulchem.blogspot.com/2010/03/reaction-attempts-on-chemspider.html
14. SMILES Wikipedia Entry http://en.wikipedia.org/wiki/Simplified_molecular_input_line_entry_specification
15. ChemSpider Web Site http://www.chemspider.com/
16. UC archive Drexel server (ZIP) http://showme.physics.drexel.edu/usefulchem/archives/usefulchem2010-04-27.zip
17. UC archive on lulu.com (DVD) http://www.lulu.com/product/dvd/usefulchem-archive/10791847
18. UC interactive hosted format http://showme.physics.drexel.edu/usefulchem/archives/usefulchem2010-04-27/All%20Reactions.html
19. Bradley, J.-C.; Lang, A.. Reaction Attempts Reactants and Products. UsefulChem. 2010-04-27.

URL:http://spreadsheets.google.com/pub?key=toMz3Kp3T2EAUDF3MSGCquw&output=xls. Accessed: 2010-04-27.

(Archived by WebCite® at http://www.webcitation.org/5pIsFEbT9)
20. Bradley, J.-C.; Lang, A.. Reaction Attempts RXIDs. UsefulChem. 2010-04-27.

URL:http://spreadsheets.google.com/pub?key=teD5QLZ7shgvDBSNp8DnonQ&output=xls. Accessed: 2010-04-27.

(Archived by WebCite® at http://www.webcitation.org/5pIs2eh62)

ONS Books Wiki

I recently reported on our use of Nature Precedings to archive different editions of the ONS Solubility Challenge book. One of the advantages is that Precedings automatically alerts visitors if more recent editions exist.

However, today I learned that there is a glitch to this system: it is not possible to link individual versions on Precedings to a corresponding book edition on LuLu. That means that if you find yourself on the Nature Precedings entry and want to order the book from LuLu it isn't obvious at all how to do so.

To resolve this issue once and for all I just created a wiki page (ONSbooks.wikispaces.com) to track every edition of the book. This is actually better because I can also provide links to all the available data archives and blog posts corresponding to each edition.

This is also the page where we will keep track of every edition of other Open Notebook Science books. The next one to be published shortly is for the UsefulChem project.

Scientists Embrace Openness Article in Science Careers

Chelsea Wald just published an article in Science Careers: Scientists Embrace Openness (April 9, 2010). She interviewed several people in the Open Science movement including Jonathan Eisen, Steve Koch, Anthony Salvagno, Carl Boettiger and myself.

The article covers Open Notebook Science, Open Data and associated themes. I think it presents a view of the most commonly discussed advantages and disadvantages very well.

One section was particularly relevant to an issue I recently posted about - (and discussed on FriendFeed):

Open Notebook Science advocates claim that being open may protect a scientist's ideas rather than exposing them to theft. Newton's decision to conceal his findings within an anagram made it harder for him to prove priority over rival Gottfried Leibniz. Open Notebook scientists say all they need to do is point to their open notebooks to show that they had an idea or found a result first.

ONS t-shirts from Zazzle

Inspired by Graham Steel, I just received my t-shirt with an Open Notebook Science Logo and a picture of our crystal on the cover of our ONS Solubility Challenge book.

I was going to set up an ONS store but Zazzle does not permit zero royalties (don't see the logic there). But making up t-shirts on Zazzle is super simple - just grab a logo of your choice from the ONSclaims wiki.

Any other pic is your choice - this is the crystal from UCEXP150C


You can also order all kinds of other personalized things, including coffee cups.

Bipolar Electrodeposition of CdS: Scientific Results in Limbo?

There has been a lot of discussion about the fear of getting "scooped" as a reason to be weary of using new scientific publication vehicles.

These conversations can be somewhat frustrating since people don't necessarily use the same definition of that term. Even dictionaries don't use the same language. For example, Dictionary.com has:

to get the better of (other publications, newscasters, etc.) by obtaining and publishing or broadcasting a news item, report, or story first: They scooped all the other dailies with the story of the election fraud.

Wiktionary has:

To learn something, especially something worthy of a news article, before (someone else). The paper across town scooped them on the City Hall scandal.

Depending on the definition used, one could argue that in the story I'm going to tell I got scooped or I did the scooping. Some people use the term to imply that a malicious act has taken place. The classic scenario is that one would blog about their research and a nefarious individual would appropriate their results and submit as their own for publication in a peer reviewed journal.

That isn't scooping - it is fraud - and I want to be clear that this is not what I am suggesting happened here.

Two months ago I was asked to review an article for the ACS journal Langmuir. Before 2005 one of my main research areas was bipolar electrodeposition and so I still get asked periodically to review papers in that field.

Not only was this paper in that area but it reported on exactly the same experimental design we had previously reported: the bipolar electrodeposition of cadmium sulfide. The solvent, reagents and substrate were different but it was the same material made by the same process.

Although 2 of our papers were listed in the references the key report was not. I noted this in my review but I was surprised to find that the paper was published without that correction. I contacted the editor of Langmuir to find out what happened. I thought perhaps the authors disagreed with some technical issue in our report.

But what actually happened is that the authors requested that the reference be included in the supplementary data section instead of the regular reference section of the article because it was not a peer-reviewed article. The editor thought that was a reasonable request and complied.

I was quite surprised by this because Langmuir - or ACS journals in general - do not have a formal policy on requiring references to be peer-reviewed. In fact, a quick search for "unpublished results" on Langmuir reveals many articles which use that as an acceptable reference.

I could understand not wanting to cite a blog post with unsubstantiated claims but the document in question is very thorough - it includes a systematic review of prior art, detailed experimental description and characterization data.

This is actually an example of a "SMIRP Knowledge Product". It is a publication device that I used to make public single experimental results from work that was recorded by my research group in the SMIRP Knowledge Management System we used at the time as a laboratory notebook.

The system was built on interlinked modules designed to produce "Knowledge Products" based on a combination of manual and automated workflows. For example, the module generating reviews of prior art was based on "Knowledge Filters" uncovering the novelty of the experiment in question by filtering precedents for relevant aspects.

In the case of "Bipolar Electrodeposition of Cadmium Sulfide onto a Tip of a Carbon Nanotube", the relevant knowledge filters were "Bipolar Electrodeposition", "Electrodeposition onto Carbon Nanotubes" and "Electrodeposition Approaches to Synthesize Cadmium Sulfide". Other modules generated the experimental description, results, discussion, conclusion and reference sections. In this way, not only could the experiment be fully documented but its context within the field could be extremely well defined in a systematic way.

With a workflow to create these knowledge products we still needed a way to communicate them. Back in 2003 options were far more limited than they are now. But luckily (or so we thought) at this time Elsevier was running the Chemistry Preprint Server. They offered a place to upload documents such as these and provided a way of citing them. We used the recommended citation format aggressively, including peer-reviewed articles such as this one from Springer.

However, attempting to access these documents today using the official links gives this as a result:
In what is probably one of the worst scientific publisher PR moves in recent memory, Elsevier broke all the hyperlinks they told their authors to use for citations. If you do some research you will find that the documents are still available from http://www.sciencedirect.com/preprintarchive but you have to register to even perform a search to find them! This requirement removes them from indexing by Google. Coupled with the broken citation links these documents are now very far removed from likely discovery.

The story would have ended there were it not for redundancy. I also uploaded copies to Drexel's institutional repository (DSpace), which are happily very well indexed by Google - and perhaps more critically - by Google Scholar. I had not fully appreciated the value of institutional repositories until I noticed that they are treated by some important databases as collections of scholarly works.

So what are the lessons for all the stakeholders?

For those who have scientific results that can be published as articles and MUST be published in ACS journals - send your manuscripts in. If you post them on your institutional repository first they may end up in limbo -they DO qualify as publications preventing you from submitting them as manuscripts to ACS journals - and they may NOT qualify as publications when you try to cite them in ACS journals.

But what about scientific result that cannot be published as manuscripts. The Knowledge Products are unlikely to be accepted by regular journals for several reasons. First they communicate only a single experimental result. Articles generally require narratives. Second, if some of them do get published in traditional journals, there will be copyright conflicts. The Knowledge Filters for the review of prior art will be identical for similar experiments. For example the Knowledge Product for the "Bipolar Electrodeposition of CdS on one Tip of a Carbon Nanotube" Will have identical prior art to the "Bipolar Electrodeposition of Cd on one Tip of a Carbon Nanotube" except for the section of the electrodeposition of Cd or CdS. And no - I don't think it is a good use of my time to move words around for every document to get around copyright issues.

Some of the Knowledge Products were incorporated into full articles when it made sense. But many, including the one under discussion here was not. So publishing this work as part of a full article was never even an option. There are so many scientific results like this that fall into that kind of limbo. Even today there are no really good publication vehicles for these types of results - besides institutional repositories. PLoS ONE might come to mind as an option but I don't think it fits their mandate to publish single experiments like this. And if they did it would be extremely expensive if they did not waive author fees every time. ChemSpider Synthetic Pages and similar initiatives might work for organic chemistry but this is materials science.

Considering all of these difficulties over the years is really the main motivation behind our migration away from a login based system like SMIRP to our adoption of Open Notebook Science based on a wikis and blogs, which are very efficiently indexed in real time by Google and thus easily discoverable without additional formatting work.

For publishers and authors, do you really think it is in your best interest to have a statement in the introduction about prior art say "To our knowledge, reports of bipolar electrodeposition of compounds have not been previously published." when a simple Google search shows that is not the case for the compound you are electrodepositing? I suppose the argument is that the term "published" is used with the technical interpretation of being "published under peer-review". It would have been better to at least make that explicit to avoid confusion. But the bottom line is that someone wanting to perform bipolar electrodeposition of cadmium sulfide will quickly find both reports and will learn two ways of doing it.

Beer Chemistry Quiz on ChemTiles

During the Chemical Information Retrieval course I taught in the Fall of 2009, Alex Bilinski did a project on the chemistry of beer. He created a set of images with information that is either true or false in any context.

I just added these images of beer chemistry to the ChemTiles game, which I am actively using in my current Organic Chemistry I (CHEM241) course. Andrew Lang has made it very easy to add content to the game by simply uploading to a Flickr group. The category is determined by the Flickr tag.

I won't be testing my organic students on beer but they might find it fun to play that category. Alex wrote a fascinating report "Beer Flavor Compounds and Detection Methods" that can be used as a study guide for the quiz.

My students are currently competing on the topics of Lewis structures, hybridization, Newman projections and nomenclature. They simply need to sign in by entering "contest1" for the group name. The student in my class with the highest score for the contest1 group at the end of class (10:50 AM) on April 9, 2010 will win an organic chemistry textbook. As I have done in previous classes I'll run a few of these contests over the term with increasing amounts of course content.

It turns out that the ChemTiles game is very convenient to play on a Droid phone (and presumably on an iPhone although I have not checked that yet). For many students this might be a preferred way to review material before tests when on the move. I'll find out at the end of the term.