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

    This article has 4 authors:
    1. Emanuele Loffredo
    2. Mauro Pastore
    3. Simona Cocco
    4. Rémi Monasson
    This article has been curated by 1 group:
    • 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

    This article has 5 evaluationsAppears in 1 listLatest version Latest activity
  2. The Demodifier: a tool for screening modification-induced alternate peptide taxonomy in palaeoproteomics

    This article has 1 author:
    1. Miranda Evans

    Reviewed by Peer Community in Archaeology

    This article has 1 evaluationAppears in 1 listLatest version Latest activity
  3. Unbiased whole genome comparison of Pan paniscus (bonobo) and Homo sapiens (human) through a novel sequence match-based approach

    This article has 11 authors:
    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

    This article has 3 evaluationsAppears in 1 listLatest version Latest activity
  4. Uncovering Developmental Lineages from Single-cell Data with Contrastive Poincaré Maps

    This article has 6 authors:
    1. Nithya Bhasker
    2. Hattie Chung
    3. Louis Boucherie
    4. Vladislav Kim
    5. Stefanie Speidel
    6. Melanie Weber

    Reviewed by Arcadia Science

    This article has 4 evaluationsAppears in 1 listLatest version Latest activity
  5. Identification of type 2 diabetes- and obesity-associated human β-cells using deep transfer learning

    This article has 9 authors:
    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
    This article has been curated by 1 group:
    • 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

    This article has 8 evaluationsAppears in 1 listLatest version Latest activity
  6. A Transparent and Generalizable Deep Learning Framework for Genomic Ancestry Prediction

    This article has 7 authors:
    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

    This article has 1 evaluationAppears in 1 listLatest version Latest activity
  7. Evidence of off-target probe binding in the 10x Genomics Xenium v1 Human Breast Gene Expression Panel compromises accuracy of spatial transcriptomic profiling

    This article has 4 authors:
    1. Caleb Hallinan
    2. Hyun Joo Ji
    3. Steven L Salzberg
    4. Jean Fan
    This article has been curated by 1 group:
    • 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

    This article has 5 evaluationsAppears in 1 listLatest version Latest activity
  8. ASPEN: Robust detection of allelic dynamics in single cell RNA-seq

    This article has 4 authors:
    1. Veronika Petrova
    2. Muqing Niu
    3. Thomas Vierbuchen
    4. Emily S Wong

    Reviewed by Review Commons

    This article has 4 evaluationsAppears in 1 listLatest version Latest activity
  9. Sequence Modeling Is Not Evolutionary Reasoning

    This article has 5 authors:
    1. Yasha Ektefaie
    2. Andrew Shen
    3. Lavik Jain
    4. Maha Farhat
    5. Marinka Zitnik

    Reviewed by Arcadia Science

    This article has 5 evaluationsAppears in 1 listLatest version Latest activity
  10. SeuratExtend: Streamlining Single-Cell RNA-Seq Analysis Through an Integrated and Intuitive Framework

    This article has 4 authors:
    1. Yichao Hua
    2. Linqian Weng
    3. Fang Zhao
    4. Florian Rambow

    Reviewed by GigaScience

    This article has 2 evaluationsAppears in 1 listLatest version Latest activity
Page 1 of 125 Next