Publications

publications by categories in reversed chronological order. generated by jekyll-scholar.

2026

  1. DrugDiscoveryBench: Can Coding Agents Assist Early-Stage Drug Discovery?
    Afra Feyza Akyürek*, Xinming Tu*, Alec Gutmanstein, Jason Qin, Sofia Monasdotter, 15 more authors, and Sami Hassaan
    Scale Labs preprint 2026·Code·Supp·Blog·Website
    Benchmark of 82 expert-curated, verifiable early-stage drug-discovery tasks for evaluating coding agents across target identification, patent mining, structure-activity analysis, and lead optimization.
  2. BiomniBench: Process-level Evaluation of LLM Agents for Real-world Biomedical Research
    Yuanhao Qu*, Yingzhou (Minta) Lu*, Xinming Tu, Serena Zhang, Tianwei She, 18 more authors, and Kexin Huang
    bioRxiv 2026
  3. BenchGuard: Who Guards the Benchmarks? Automated Auditing of LLM Agent Benchmarks
    Xinming Tu, Tianze Wang, Yingzhou (Minta) Lu, Kexin Huang, Yuanhao Qu, and Sara Mostafavi
    COLM 2026
  4. A modality gap in personal-genome prediction by sequence-to-function models
    Xinming Tu, Anna Spiro, Maria Chikina, and Sara Mostafavi
    bioRxiv 2026
  5. A scalable approach to investigating sequence-to-expression prediction from personal genomes
    Anna E Spiro*, Xinming Tu*, Yilun Sheng, Alexander Sasse, Rezwan Hosseini, 1 more author, and Sara Mostafavi
    Nature Methods 2026
  6. Deep genomic models of allele-specific measurements
    Xinming Tu*, Alexander Sasse*, Kaitavjeet Chowdhary*, Anna E Spiro, Liang Yang, 2 more authors, and Sara Mostafavi
    RECOMB 2026

2025

  1. Contrasting genetic predisposition and diagnosis in psychiatric disorders: A multi-omic single-nucleus analysis of the human OFC
    Nathalie Gerstner, Anna S Fröhlich, Natalie Matosin, Miriam Gagliardi, Cristiana Cruceanu, Maik Ködel, Monika Rex-Haffner, Xinming Tu, 3 more authors, and Janine Knauer-Arloth
    Science Advances 2025
  2. GAME: Genomic API for Model Evaluation
    Ishika Luthra, Satyam Priyadarshi, Rui Guo, Lukas Mahieu, and  others
    bioRxiv 2025
    Large-scale community-led model benchmarking system

2024

  1. A Supervised Contrastive Framework for Learning Disentangled Representations of Cell Perturbation Data
    Xinming Tu, Jan-Christian Hütter, Zitong Jerry Wang, Takamasa Kudo, Aviv Regev, and Romain Lopez
    MLCB 2024

2023

  1. What Should Data Science Education Do with Large Language Models?
    Xinming Tu, James Zou, Weijie J. Su†, and Linjun Zhang†
    Harvard Data Science Review 2023
  2. Spatial-linked alignment tool (SLAT) for aligning heterogenous slices
    Chen-Rui Xia, Zhi-Jie Cao, Xinming Tu, and Ge Gao
    Nature Communications 2023

2022

  1. Cross-Linked Unified Embedding for cross-modality representation learning
    Xinming Tu*, Zhi-Jie Cao*, Chen-Rui Xia, Sara Mostafavi†, and Ge Gao†
    NeurIPS (Oral) 2022

2021

  1. Identifying complex motifs in massive omics data with a variable-convolutional layer in deep neural network
    Jing-Yi Li*, Shen Jin*, Xinming Tu, Yang Ding†, and Ge Gao†
    Briefings in Bioinformatics 2021

2019

  1. Expectation pooling: an effective and interpretable pooling method for predicting DNA–protein binding
    Xiao Luo*, Xinming Tu*, Yang Ding, Ge Gao†, and Minghua Deng†
    Bioinformatics 2019
  2. An exact transformation of convolutional kernels enables accurate identification of sequence motifs
    Yang Ding, Jing-Yi Li, Meng Wang, Xinming Tu, and Ge Gao
    bioRxiv 2019