Back

Illuminating the Druggable Genome through Patent Bioactivity Data

Leach, A. R.; Magarinos, M. P.; Gaulton, A.; Felix, E.; Kiziloren, T.; Arcila, R.; Oprea, T. I.

2022-07-18 bioinformatics
10.1101/2022.07.15.500187 bioRxiv
Show abstract

The patent literature is a potentially valuable source of bioactivity data. The SureChEMBL database (https://www.surechembl.org/) is a publicly available large-scale resource that contains compounds extracted on a daily basis from the full text, images and attachments of patent documents, through an automated text and image-mining pipeline. In this paper we describe a process to prioritise 3.7 million life science relevant patents obtained from SureChEMBL, according to how likely they were to contain bioactivity data for potent small molecules on less-studied targets, according to the classification developed by the Illuminating the Druggable Genome (IDG) project. The overall goal was to select a smaller number of patents that could be manually curated and incorporated into the ChEMBL database. We describe the approach taken, the results obtained, and provide some illustrative examples.

Matching journals

The top 9 journals account for 50% of the predicted probability mass.

1
Scientific Reports
○ 3612 papers in training set
Top 2%
12.9%
Similar papers in this journal
2
Scientific Data
○ 209 papers in training set
Top 0.3%
8.0%
Similar papers in this journal
3
Computational and Structural Biotechnology Journal
● 242 papers in training set
Top 0.4%
6.4%
Similar papers in this journal
4
Briefings in Bioinformatics
◐ 354 papers in training set
Top 1%
5.6%
Similar papers in this journal
5
iScience
○ 1154 papers in training set
Top 3%
4.5%
Similar papers in this journal
6
International Journal of Molecular Sciences
○ 494 papers in training set
Top 2%
4.4%
7
Artificial Intelligence in the Life Sciences
○ 13 papers in training set
Top 0.1%
3.5%
8
Journal of Cheminformatics
○ 29 papers in training set
Top 0.2%
3.5%
9
PLOS ONE
● 5266 papers in training set
Top 35%
3.3%
50% of probability mass above
10
Journal of Chemical Information and Modeling
● 238 papers in training set
Top 1%
2.8%
11
RSC Advances
● 22 papers in training set
Top 0.2%
2.7%
12
Molecules
○ 39 papers in training set
Top 0.4%
2.5%
13
Journal of Medicinal Chemistry
● 77 papers in training set
Top 0.4%
2.4%
14
ACS Omega
● 105 papers in training set
Top 1%
1.8%
15
Frontiers in Pharmacology
○ 111 papers in training set
Top 1%
1.8%
16
Communications Chemistry
○ 48 papers in training set
Top 0.5%
1.7%
17
mAbs
○ 32 papers in training set
Top 0.3%
1.5%
18
SLAS Discovery
○ 25 papers in training set
Top 0.2%
1.2%
19
Database
◐ 61 papers in training set
Top 0.6%
1.2%
20
Nature Protocols
○ 33 papers in training set
Top 0.4%
1.2%
21
European Journal of Pharmacology
○ 15 papers in training set
Top 0.3%
1.2%
22
European Journal of Medicinal Chemistry
○ 17 papers in training set
Top 0.2%
1.2%
23
Pharmaceuticals
○ 34 papers in training set
Top 0.7%
1.2%
24
Chemical Science
● 73 papers in training set
Top 1%
1.1%
25
Bioinformatics Advances
◐ 203 papers in training set
Top 4%
1.1%
26
ACS Pharmacology & Translational Science
● 40 papers in training set
Top 0.7%
0.9%
27
The Pharmacogenomics Journal
○ 11 papers in training set
Top 0.2%
0.9%
28
PLOS Computational Biology
● 1863 papers in training set
Top 19%
0.9%
29
Bioinformatics
◐ 1204 papers in training set
Top 9%
0.6%
30
Nature Communications
○ 5641 papers in training set
Top 58%
0.6%

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.