Bioinformatics

Predicting functional constraints across evolutionary timescales with phylogeny-informed genomic language models

1. Chengzhong Ye
2. Gonzalo Benegas
3. Carlos Albors
4. Jianan Canal Li
5. Sebastian Prillo
6. Peter D. Fields
7. Brian Clarke
8. Yun S. Song

Reviewed by Arcadia Science

Product-stabilized filamentation by human glutamine synthetase allosterically tunes metabolic activity

1. Eric Greene
2. Richard Muniz
3. Hiroki Yamamura
4. Samuel E Hoff
5. Priyanka Bajaj
6. D John Lee
7. Erin M Thompson
8. Angelika Arada
9. Gyun Min Lee
10. Massimiliano Bonomi
11. Justin M Kollman
12. James S Fraser

• Curated by eLife

  eLife Assessment

  This manuscript employs cryo-EM, mutational analysis, and biochemical assays to explore the molecular basis by which glutamine promotes filamentation and regulates the activity of human glutamine synthetase (hGS) by stabilizing interactions between hGS decamers. The studies supporting this mechanism are solid, but could be improved by providing more clarity and by addressing methodological issues in the cryoEM data processing workflow. This work will be of particular interest and useful to groups interested in understanding the molecular basis of nutrient sensing, cellular metabolism, and structural regulation of enzyme activity.

Reviewed by eLife

Macroscopic Analyses of RNA-Seq Data to Reveal Chromatin Modifications in Aging and Disease

1. Achal Mahajan
2. Francesca Ratti
3. Ban Wang
4. Hana El-Samad
5. James H Kaufman
6. Vishrawas Gopalakrishnan

• Curated by eLife

  eLife Assessment

  This valuable study analyzes aging-related chromatin changes through the lens of intra-chromosomal gene correlation length, which is a novel computational metric that captures spatial correlations in gene expression along the chromosome. The authors propose that this metric reflects chromatin structure and can serve as a proxy for its changes during aging. While currently the strength of evidence is somewhat incomplete, if revised with further supporting data, this work will provide a systems-level understanding of aging and genome regulation, which is predicted to have a substantive impact on the field.

Reviewed by eLife

Restoring data balance via generative models of T cell receptors for antigen-binding prediction

1. Emanuele Loffredo
2. Mauro Pastore
3. Simona Cocco
4. Rémi Monasson

• Curated by eLife

  eLife Assessment

  This valuable study introduces a data augmentation approach based on generative unsupervised models to address data imbalance in immune receptor modeling. Support for the findings is solid, showing that the use of generated data increases the performance of downstream supervised prediction tasks, e.g., TCR-peptide interaction prediction. However, the validation, mainly relying on synthetic data, could be completed, especially regarding unseen epitopes, and given the exclusive focus on CDR3β. The results should be of interest to the communities working on immunology and biological sequence data analysis.

Reviewed by eLife

The Demodifier: a tool for screening modification-induced alternate peptide taxonomy in palaeoproteomics

1. Miranda Evans

Reviewed by Peer Community in Archaeology

Unbiased whole genome comparison of Pan paniscus (bonobo) and Homo sapiens (human) through a novel sequence match-based approach

1. Christof Rosler
2. Derick DuFriend
3. Evonn Annor
4. Rita Njoroge
5. Elena Davis
6. Ywahae Law Anthem
7. Celestino Velásquez
8. William P. Ranahan
9. Matthew H. Goelzer
10. Stephen Wheat
11. Julianna A. Goelzer

Reviewed by Arcadia Science

Uncovering Developmental Lineages from Single-cell Data with Contrastive Poincaré Maps

1. Nithya Bhasker
2. Hattie Chung
3. Louis Boucherie
4. Vladislav Kim
5. Stefanie Speidel
6. Melanie Weber

Reviewed by Arcadia Science

Identification of type 2 diabetes- and obesity-associated human β-cells using deep transfer learning

1. Gitanjali Roy
2. Rameesha Syed
3. Olivia Lazaro
4. Sylvia Robertson
5. Sean D McCabe
6. Daniela Rodriguez
7. Alex M Mawla
8. Travis S Johnson
9. Michael A Kalwat

• Curated by eLife

  eLife assessment

  This is a useful study that applies deep transfer learning to assign patient-level disease attributes to single cells of T2D and non-diabetic patients, including obese patients. This analysis identified a single cluster of T2D-associated β-cells; and two subpopulations of obese- β-cells derived from either non-diabetic or T2D donors. The findings were validated at the protein level using immunohistochemistry on islets derived from non-diabetic and T2D organ donors, contributing solid experimental evidence for the computational analyses.

Reviewed by eLife

A Transparent and Generalizable Deep Learning Framework for Genomic Ancestry Prediction

1. Camille Rochefort-Boulanger
2. Matthew Scicluna
3. Raphaël Poujol
4. Jean-Christophe Grenier
5. Pierre Luc Carrier
6. Sébastien Lemieux
7. Julie G Hussin

Reviewed by Arcadia Science

Evidence of off-target probe binding in the 10x Genomics Xenium v1 Human Breast Gene Expression Panel compromises accuracy of spatial transcriptomic profiling

1. Caleb Hallinan
2. Hyun Joo Ji
3. Steven L Salzberg
4. Jean Fan

• Curated by eLife

  eLife Assessment

  This valuable study identifies and characterizes probe binding errors in a widely used commercial platform for visualizing gene activity in tissue samples, discovering that at least 21 out of 280 genes in a human breast cancer panel are not accurately detected. The authors provide convincing evidence for their findings validated against multiple independent sequencing technologies and reference datasets. Given the broad adoption of this spatial gene detection platform in biomedical research, this work provides an essential quality control resource that will improve data interpretation across numerous studies.

Reviewed by eLife