Depth-First

Why Organic Chemistry is Like Facebook

2012-05-21T00:00:00-07:00

Last week’s news was dominated by the initial public offering of Facebook. Although it’s unlikely that many investors understood exactly what they were buying, much of the company’s value derives from the massive social graph created within the last few years, and tools for exploiting it commercially.

Graphs: Social and Chemical

One of the most practical (and profitable) ways to think of organic chemistry is from the perspective of functional group transformations. One functional group can be converted to another using one or more reagents under a given set of conditions. This is represented below, where “Fg” is a functional group and the arrow represents a single reaction:

Just as Facebook only became interesting after many people had joined and friended each other, organic chemistry only becomes interesting after many functional groups are connected to each other through reactions:

This representation is called a graph, where the circles represent “nodes” and the arrows represent “edges.” Many real-world situations can be modeled as graphs, including both Facebook and organic synthesis.

The analyses applied to a graph modeling one situation can often be usefully applied to graphs modeling a completely different situation. For example, a listing of all the people your friends know and what they like can help you discover people and interests you didn’t know about before. Likewise, a listing all of the transformation a functional group can undergo can help you find new substances easily accessible from substances you already have.

Reaction Maps

One of the most familiar kinds of graph is a roadmap linking destinations through roads. For this reason, the idea of linking functional groups through reactions is also called a “reaction map.” Reaction maps are usually discussed in the context of teaching and learning organic chemistry, but they are not at all limited to this use. For more, see:

Whether using the term “graph” or “map” to describe the model that results from connecting known functional groups through known reactions, the concepts remain the same.

A GPS for Organic Synthesis

Although the value of a network increases with the number of nodes, so does the difficulty of navigation. The nearly 200 years during which chemists have been building the organic synthesis graph have seen the development of a very complex web of knowledge that can be utterly confusing without the right tools.

As part of a collaboration with James Ashenhurst, I’ve been looking at how to make organic synthesis more accessible to students. The tool we’ve developed, Reagents, is an app that runs on iPhone and iPad devices.

Until recently, Reagents was simply a listing of organic synthesis reagents. Although useful in its own right, we thought we could do better.

That’s why the most recent version of Reagents introduced two powerful new features: “Synthesis” and “Retrosynthesis”. Given any of the functional groups commonly used in Organic 1/2 level courses, it’s now possible to find reactions that the group will participate in (“Synthesis”) or which will generate the functional group (“Retrosynthesis”). From there, the necessary reagents can be found.

Exposing this view of the organic synthesis graph was not conceptually difficult, but did require quite a few changes to both the Reagents source code and data format.

Conclusions

When viewed as a graph or map, organic chemistry can become become both very approachable and very useful.

The key is using the right tools. Up until recently, books were the primary means for researchers and students alike to navigate the organic synthesis graph. As a comparison between Facebook and the old White Pages should make clear, books provide a suboptimal user interface to graph data.

New devices like iPad offer fundamentally new ways of learning and using the organic synthesis graph. Reagents offers a new perspective on this old problem.

Balancing Chemical Equations in ReactionMate Part 1: More Than Meets the Eye

2012-05-17T00:00:00-07:00

Balancing chemical equations is one of the first things taught in most introductory chemistry classes. As such, it might seem as if everything there is to know about the topic was published 75 years ago or more by chemists long since gone.

To my surprise, I found this was simply not so.

This post, the first in a series, discusses why a cheminformatics fan like myself would even consider the problem of balancing chemical equations as a serious topic, and what a deeper understanding of the subject might mean for you.

Origins: ReactionMate

ReactionMate (App Store) at the moment is a little iOS app with big ambitions. It aims to become a useful companion to anyone studying, performing, or using chemical reactions.

One of the fundamental requirements for processing chemical reactions computationally is the automatic balancing of chemical equations, a function ReactionMate performs relatively well.

Given this foundation, there are many interesting directions the app could be taken, both in education and research. Hopefully, I’ll be discussing some of them later.

When I started working on ReactionMate, my main motivation was to better learn iOS development with Objective-C. A reaction balancer, so I thought, would be an easy project.

Although I did reach the first milestone and now have an app for sale to show for the effort, the path was much longer and darker than I imagined.

Balancing By Inspection

You may remember back to learning how to balance chemical equations in General Chemistry. A common method still taught is known as balancing by inspection. This explanation from Khan Academy is typical:

Although a fine approach for balancing simple equations by hand, balancing by inspection does not lend itself well to use in a software algorithm. I suspect that the same thing that makes balancing by inspection so difficult for students to master is the same thing that makes it unsuitable for use as algorithm: it’s hit or miss and then try again.

Balancing With Matricies/Linear Algebra

A more general and deterministic method for balancing chemical equations uses linear algebra. Given the constraint that the number of atoms of a given type on the left of the equation must equal the number of that type of atom on the right, it’s possible to set up a system of linear equations and use matrix operations to solve for coefficients. The video below shows how this works conceptually:

Issues

Implementing the necessary matrix algebra in Objective-C was not terribly difficult in a first pass. Coefficients for the first few chemical equations were found without problems.

However, as the number and complexity of tested equations grew, problems began to surface - each demanding a solution. The next installment will discuss some of the issues and how they were addressed.

Conclusions

Balancing chemical equations is a deep topic at the intersection between mathematics and chemistry. Balanced chemical equations also lie at the heart of many important areas in chemistry. With academic papers appearing as recently as a couple of years ago, this topic is full of surprises.

Why ACS Must Come Clean on Journal Publication Costs

2012-03-14T00:00:00-07:00

The now-fading discussion of Elsevier and the Research Works Act got me thinking about the American Chemical Society, which runs a very large scientific publication business of its own. I started wondering why the ACS exists in the first place and what its long-term vision might be.

In 1938, a U.S. Congressional Charter was granted to the American Chemical Society. Although essentially honorific, this document makes for interesting reading. Front and center is Section 2, containing these inspiring words:

That the objects of the incorporation shall be to encourage in the broadest and most liberal manner the advancement of chemistry in all its branches; the promotion of research in chemical science and industry; the improvement of the qualifications and usefulness of chemists through high standards of professional ethics, education, and attainments; the increase and diffusion of chemical knowledge; and by its meetings, professional contacts, reports, papers, discussions, and publications, to promote scientific interests and inquiry, thereby fostering public welfare and education, aiding the development of our country’s industries, and adding to the material prosperity and happiness of our people.

As a longstanding member of the ACS, I question the policies ACS has pursued in light of the above statement. ACS has consistently maintained that the only way it can continue to publish journal content is to either compel authors to pay anywhere from $1,000 to $3,000 for its opt-in Author Choice Option, or to compel them to transfer all rights to the content they’re publishing so that ACS can place keep it behind a variety of paywall mechanisms.

Under either option, this is a pay to play system. There are justifiable costs that everyone must pay, and then there’s profiteering. The latter is diametrically opposed to encouraging “in the broadest and most liberal manner the advancement of chemistry”.

Having been in business for a number of years now, I fully appreciate the need to cover costs and turn a profit. But ACS is a non-profit organization. In theory, it just needs to cover costs.

What are the costs to produce the various ACS journals? Nobody asked such a question back when journals were printed on dead trees and filled library shelves. Desktop publishing, Moore’s law, and the Web have changed all of that.

Conspicuously absent from the various ACS policy statements on open access is any form of financial transparency. Must we take at face value claims that ACS publications all need to continue with business as usual lest they cease to be financially viable?

As a previous Depth-First article pointed out, the best we can do is speculate. An annual financial statement is released by ACS and it is possible to connect some dots, but this makes for poor decision-making.

Release of a detailed breakdown of the costs to produce each ACS journal would go a long way to elevating the open access debate, and could turn out to support the ACS case.

Sadly, I fear no releases of this kind of financial data will be forthcoming. Less than 10% of the total ACS budget comes from dues and meeting registration fees. ACS is clearly using the proceeds of its publication business (and Chemical Abstracts) to fund scholarships, outreach programs, webinars, social networking experiments, job fairs, lobbying efforts, scientific awards, employment surveys, and executive compensation packages, among other things.

Some of these programs offer real value. Others are counterproductive, to put it kindly. Supporting all of these activities is not the issue. The issue is to what extent ACS is paying for all of this stuff by creating and perpetuating a dysfunctional publication system that hurts chemistry in the long run.

These decisions are not unilaterally made by Madeleine Jacobs or any of the other folks whose smiling images regularly grace the pages of C&ENews. The ACS is run by its members - at least one can hope so.

If you’re an ACS member concerned with the direction ACS is headed, you have the obligation and right to ask for financial transparency around ACS publications.

Education of a Scientist

2012-02-07T00:00:00-08:00

Creating this video was fun, and there’s an obvious business model behind the company making it possible. Guess what? I just published it for free to a worldwide audience.

Why doesn’t science work this way?

George Whitesides: The Concept of the Scientific Paper is Eroding Before Our Very Eyes

2012-01-28T00:00:00-08:00

George Whitesides is well known as an innovator and one of chemistry’s most visible representatives. He has the distinction of being credited with the highest Hirsch index for any living chemist. You could say he knows something about scientific publication.

The above video excerpt was taken from an extended interview with Whitesides on the topic of publication. Although the full interview series is worth watching, most striking are Whitesides’ views on the changes happening in scientific publication and what the future is likely to hold:

One of the troubles with universities is there’s a tendency to do terrific research, embed it in prose that is impenetrable even to experts, bury it in papers, and have to everyone’s surprise nothing come out of it. … It may well be that what we have in the future is some combination of very short snips in one or another kind of extended abstract leading to, through links - through something else, more and more levels of detail.

The scientific paper evolved under an environment in which scarcity ruled. This scarcity enforced a uniform format on scientific discourse. Due to physical constraints on printing and distribution, only certain kinds of research could be published. A great deal of valuable information and effort have been wasted in the process.

The Web and desktop publishing have abolished these constraints, but most scientist continue to act as if nothing has changed. They continue to huddle around the dying embers imprimatur left over from a handful of pre-digital journals. Or they try and try again to push these journals into distribution models (Open Access) that are simply incompatible with the top-heavy organizations that have grown around outdated publication models.

Those scientist and nimble information service providers who understand that the old rules no longer apply will enjoy significant advantages that will amaze (and possibly leave behind) the ones who can’t (or won’t) adapt.

Five Things to Do Instead of Protesting the Research Works Act (HR 3699)

2012-01-23T00:00:00-08:00

The Research Works Act (‘RWA’, or HR 3699) would reverse the US federal government policy of requiring recipients of National Institutes of Health (NIH) funds to deposit copies of their research papers into PubMed Central. It would also prevent the adoption of similar policies by other federal funding agencies going forward.

As a scientist who has participated in the authoring and review of a few scientific papers in closed, for-profit journals, I believe the Research Works Act should be allowed to pass, and that opposition to it focusses on the wrong problem, despite good intentions.

Fighting The Research Works Act Is Not Worthy of Your Time or Intellect

For those still willing to entertain an alternative perspective on this charged issue, consider:

The NIH Public Access Policy in no way changes the copyright status of the works appearing on PubMed Central. Redistribution, duplication, or repurposing of works from PubMed Central could still make you liable to the usual copyright infringement penalties - even if you are the author.
The NIH Public Access Policy is assembling an incomplete corpus of scientific works. Authors not supported by NIH, which includes most in the chemical and pharmaceutical industries, and foreign authors - among others, are not subject to the NIH policy. Their works will generally never appear in PubMed Central. Although in certain situations having access to an incomplete corpus of scientific papers can be helpful, for the most part the practical utility is about on par with an Internet in which 4 out of 5 sites are blacked out.
The Public Access Policy attempts to do through legislation what scientists should be doing for themselves - namely, completing the transition from pre-digital era of publication scarcity to the post-digital era of limitless, cheap publication capacity.

Fighting the Research Works Act is one of the least effective things a scientist could do to fix a deeply dysfunctional system of scientific publication. If you must place a “Stop RWA” banner on your Twitter avatar, or write your representatives about the “Evil Scientific Publisher Lobby” to feel better, by all means do it.

You’ll accomplish little for science, however.

If on the other hand you want to pass onto the next generation a system of scientific communication that accelerates science rather than holding it back, you’ll have to work much harder and take some rather unpleasant risks.

Government can’t do it for us. It’s our mess and we need to clean it up.

Five Ways to Effect Real Change

Identify all journals in your field that: (1) require authors to transfer copyright as a condition of publication; and (2) publish any research works under terms that don’t allow free redistribution and commercial repurposing of content.
Submit no further manuscripts to the journals you identified in Step (1). Refuse all requests to review papers for these journals. Write letters to the editors of these journals explaining your actions. Publicize your actions through some public medium (e.g., blog or letter to the editor of a trade magazine).
You’ll still need to publish, of course. Find at least one journal in your field that allows you to retain copyright to your work and which makes its content available for redistribution and commercial repurposing. From now on, only submit articles to that journal and accept all reasonable requests to review papers from them.
Identify the ten most influential scientists in your field. Find out which of them publish mainly or exclusively in the journals you identified in Step (1). Write a letter to each leader explaining the crisis in scientific publication, the harm they’re doing to science and their group in continuing to publish in these journals, and the steps you’ve taken to solve the problem. Ask for their commitment to do the same as you have.
You’re very likely to get either no response from the leaders in your field or a negative one. If it should happen that you get a favorable response, ask this leader to publish an open letter to the editor of the appropriate journals explaining why their policies are detrimental to scientific progress.

Conclusions

The resolution to the scientific publishing crisis will not come through a government bailout in the form of public access policies. It will come from starving entrenched old-guard publishers of the only value they’re currently adding to the scientific publication system - imprimatur. Regardless of whether you’re a leader in your field or just a concerned scientist, imprimatur comes from the combined perceptions of you and your peers. Fortunately, perceptions can change.

The way to create a scientific publication system that advances the cause of science is to make it repugnant, ridiculous, and lonely to participate in one that doesn’t.