Biology subjects

Barnier, J.

Publications and source records attributed to Barnier, J..

2 recordsLinked to original sources

bioRxiv2025

pylifemap: Mapping Data onto the Tree of Life

O_LIThe need to visualize data associated with NCBI Taxonomy Identifiers is growing in various biological fields ranging from comparative genomics to metagenomics and metabarcoding, and even for outreach. C_LIO_LINo tool today allows visualization of such data while still keeping the full vision of the whole taxonomy, possibly causing a biased view of the data at hand. C_LIO_LIHere we introduce pylifemap, a Python package that allows users to map their own data directly on the interactive taxonomic tree proposed by Lifemap. Through a simple syntax, layers (points, lines, icons, heatmaps, etc.) depicting any type of data are superimposed on the Lifemap basemap, producing an interactive map to inspect biological datasets. The produced visualizations can easily be shared with others through notebooks or standalone HTML files, or exported as static images. C_LIO_LIWe illustrate the utility of pylifemap in the exploration of two contrasting datasets: the IUCN Red List of Threatened Species and the output of a large-scale environmental metagenomics experiment. C_LI Data/Code for peer reviewpylifemap is available at PyPI and easily installable with pip or uv. The development version can be found at https://lifemap-tol.github.io/pylifemap/ along with extensive documentation and examples.

evolutionary biology↗

bioRxiv2025

Characterization of selective pressures acting on protein sites with Deep Learning

It is often useful, in the field of molecular evolution, to identify the selective pressures acting on a particular site of a protein to better understand its function. This is typically done with likelihood-based approaches applied to codon sequences in a phylogenetic context. However, these approaches are computationally costly. Here we adapt a linear transformer neural network architecture, which has been shown to be able to reconstruct accurate pairwise distances from sequence alignments, to identify selective pressures acting on individual amino acid sites. We design different versions of the architecture and train and test them on simulations. We compare the results of one of our best models to state-of-the-art likelihood-based methods and find that it outperforms it when it is applied to data that resemble its training data, but that it performs less well when applied to datasets that do not resemble the ones the model has been trained on. In all cases, our approach operates at a fraction of the computational cost of likelihood-based methods. These results suggest that such a neural network architecture can compare very favorably to state-of-the-art approaches to characterize selection pressures acting on coding sequences, but that it must be trained on datasets representative of empirical data.

bioinformatics↗