new: "Investigating the Molecular Processes behind the Cell-Specific Toxicity Response to Titanium Dioxide Nanobelts"

  Part of Figure 2. Pathway-gene (gray)
networks of altered toxicity-related
pathways. The colors are GO categories
(yellow: apoptosis, blue: inflammation,
purple: DNA damage, red:
oxidative stress).

Omics analysis gives results for many entities at the same time. Pathways databases describe the interactions of those entities. Combining this gives a lot of statements about that state of the biological system, but without filtering looks like a hairbal, all in an attempt to see the forest for the trees. But the forest is dense, large, and full of ever moving ents. Over the years many approaches have been invented to help us see the paths through the hairball.

Describing the trees and ents The Gene Ontology (GO) and databases like Ensembl, and OMIM all work hard to describe the trees and ents in as much detail as possible. Wikidata acts as a central node bringing all the descriptions together. GO provides a common language to describe the genes with, for example, their molecular function.

Paths and maps For how the trees for a tree, we have interaction and pathway databases, like WikiPathways. These describe how the ents in the forest biological come together and work together to react to some event. That event can be the exposure to a chemical compound, like a toxicant, a drug, a nutrient, or a nanomaterial with mRNA.

Laurent Winckers and Martina Kutmon decided to test the idea of filtering the hairball with GO categories (doi:10.3390/ijms22179432). Figure 2 (see inset) shows that pathway-gene networks that we could already make can be simplified by just looking at certain GO categories and color nodes according to their biological role. A notebook was written to allow easy reproducing the results and reuse it for other datasets, other GO categories, or updated pathway models. Congrats to both with this nice paper!

What now? Of course, now that we know that this works, we have a new tool to analyze omics data and make sense of the forest. Most literature still looks at the threes. In fact, most discussion and conclusion sections in literature still try to describe biological results at a tree level. Even this paper does this. It is what we are used to. Our data analysis tools skills are improving. For me, there is one more essential aspect missing in our tools. This is what another PhD has been working on (paper pending).