Gaurav Diwan
banner
gddiwan.bsky.social
Gaurav Diwan
@gddiwan.bsky.social
Postdoc Bioquant | Uni Heidelberg
Molecular Evolution 🧬
Phylogenomics 🪾
Bioinformatics 🖥️
- Placing new #Biodiversity #Genomics genomes (e.g.) in this framework gives context to which gene functions already exist in that clade and what is new/unknown for the new one

Big thanks to Rob Russell, Paschalis, JC, Mu-en, @maxjtelford.bsky.social @guigolab.bsky.social John Colbourne!

10/10
September 23, 2025 at 9:53 AM
So take home messages:
- Overlaying function onto gene evolutionary history captures the specific biology of species without #morphology
- Phylogeny can be perfected & annotations can be perfected but if one fuses them together as they are, one could get deep #Evolution insights

9/10
September 23, 2025 at 9:53 AM
And when we look inside the clusters, we can check the exact genes that have the analogous functions leading to the adaptation to the new environment. N.B. these genes are not from the same HOG. So, they're not related by sequence but by #function...

8/10
September 23, 2025 at 9:53 AM
For the water to land nodes, terms inside a GO cluster were
Bone development in tetrapods 🦴
Open tracheal system in arthropods
Seed/Leaf development in plants 🍀
These were all anatomical structures which organisms needed to survive on land.

Again... these were captured from gene annotations!
7/10
September 23, 2025 at 9:53 AM
Next we analyzed the functions emerging at nodes that underwent large phenotypic transitions. In our tree, we had 3 nodes where species moved from water to land and 5 nodes where organisms became multicellular.

We enriched GO terms at transition nodes and clustered them by semantic similarity

6/10
September 23, 2025 at 9:53 AM
Eukaryotes = splicing + cell cycle + transcription/translation
Metazoa = differentiation + embryo development + nervous systems
Vertebrates = immune response + digestive systems + circulatory systems

So we are recovering characteristic #traits from gene annotations!

5/10
September 23, 2025 at 9:53 AM
Instead of focusing on which exact genes were gained or to avoid over-interpreting the ancestral reconstruction of our sets, we analysed the enriched GO terms and categories of pathways gained at these nodes. We found interesting specificity here
#genomebiology

4/10
September 23, 2025 at 9:53 AM
Nodes with the largest HOG + pathway component gains were the ancestors of:
- All Eukaryotes
- Flowering plants 🌻
- Slime molds (!)
- Metazoans 🦐
- Vertebrates 🚶‍♂️
#EvoBio #evolbio

3/10
September 23, 2025 at 9:53 AM
We sorted all genes into HOGs (at the root of the species tree) and traced the history of each HOG on the species tree. We also did the same for the orthologs of each gene.
We used:
Each HOG = one gene that duplicated/diverged
Each ortholog = KEGG/Reactome pathway component
#phylogenomics
2/10
September 23, 2025 at 9:53 AM