JSON of literature search results

Mr. George L. Malone

15^th of March, 2021

This repository contains the contents of the code executor cells in a ChemInfo view for viewing and adding to a JSON of details of significantly dysregulated metabolites under SARS-CoV-2 disease status, as found in scientific literature within a specific category or set of categories. The purpose of storing the active code here is for improved version control, and for better exposure of the developed code.

A large JSON containing data extracted from PathBank is used for adding to the literature search results. This is used to permit greater control of certain identifying features of the metabolites entered, as a previous attempt to curate some literature information lacked rigour, resulting in jumping through many hoops to attempt to find and identify correct data (see glm729/d3fdg). Thus, this should solve the problem, in a way, by starting again and eliminating one of the root problems with the originally requested project specifications.

This work is also a partial spin-off of an attempt to aggregate certain curated chemical data. Work on the aggregator is ongoing, and may be provided on GitHub in future.

Current capabilities

In its current state, the view in ChemInfo is able to display all data present in the PathBank and literature JSON files. The PathBank data are displayed in a generated HTML table (see showPathbankData.js), and the literature data are displayed in Slick Grid modules (which appear to be an implementation of this library). When clicking a row in the literature data, any information in the literature for the clicked chemical is displayed, as well as a tabular form for the user to add further data to the literature JSON if desired. If new data are added, the user is prompted (i.e. a button is displayed) to download the new/augmented file.

Below is a screenshot (at 2/3 zoom) of the view in its current state. The visual structure of the view is relatively rudimentary at this stage, e.g. module boundaries are not hidden, buttons are not properly aligned. This will likely be improved as the view is developed.

Excluded data

Two cells, for reading in the PathBank JSON and the literature JSON, are two lines each, so these have been excluded. They are literally of the form:

let data = API.getData("jsonName").resurrect();
API.createData("dataName", data);

where name is used here as a placeholder. The initial entry in the literature JSON is also excluded, as this is a bit boring and might be a bit too much clutter (but can be uploaded if of interest, it doesn't contain confidential data).

The PathBank JSON is excluded, mainly because it is 172MB, but also because I do not own the data -- the original dataset can be found here, and is the second entry in the list (Metabolite names linked to PathBank pathways CSV). A description of the reduction process from the original 1.2GB CSV to the 172MB JSON will be described in another repository, which is as yet not prepared.