MolPress is an open source WordPress plugin, based on the WebMolKit library that provides cheminformatics functionality for web applications. The project was created to address a key shortcoming: on the one hand, using popular blogging/content management systems like WordPress to implement an electronic lab notebook (ELN) system is a very attractive proposition. On the other hand, such general purpose document creation tools force the scientist-author to choose between human-readable graphical figures or the underlying machine-readable data, but not both (at least not without considerable inconvenience).

The core functionality of the plugin is to provide a bridge between the two: when documenting an experiment, the scientist provides the raw data, in its machine readable form (e.g. a molecular structure, a collection of structures & measurements, a chemical reaction, etc.). Such data is stored in a form that is not very meaningful to people – to the untrained eye it looks like gibberish. The MolPress plugin parses these raw data objects and renders them in the graphical diagram format the scientists expect to see in a publication.

For example, this is what the structure of acetamide looks like when it is stored in a concise machine-friendly representation:

SketchEl!(4,3)
C=-3.8000,3.5500;0,0,i3
C=-2.5010,4.3000;0,0,i0
N=-1.2019,3.5500;0,0,i2
O=-2.5010,5.8000;0,0,i0
1-2=1,0
2-3=1,0
2-4=2,0
!End

And this is what chemists expect to see:

The MolPress plugin translates the raw datastructure into a visual picture at the time of viewing, using the SVG format (Scalable Vector Graphics), which means that the pictures are rendered perfectly at any resolution, and also look equally perfect when printed out on paper or exported to a container format, such as PDF.

If the only thing that mattered was being able to prepare a document that makes sense to human chemists, then it would be quite sufficient to use a software package to draw the structure, turn it into a picture of some sort, and then embed that picture into the web page or print-ready document format. This is basically what most people do now. This is a problem, though, because it means that there is no straightforward way to search your document by structure, or to compose meaningful queries for the various other kinds of metadata. It is also not possible for algorithms to pull out your data (e.g. structure-to-property correlations) and feed them into models, or any other such higher order purpose. These activities can only be accomplished by external curation (which is very expensive, and is an unnecessary duplication of effort) or by data mining with various kinds of machine learning (which really doesn’t work very well).

The solution is to keep the original data, and render it too. The MolPress plugin makes this possible in a straightforward way.

Keep reading for installation instructions.