aiqbench analysis
This research introduces PT-RAG, a retrieval-augmented generation framework that enhances deep learning models for predicting cellular responses to genetic perturbations by leveraging related perturbation contexts. By adaptively selecting relevant data points across different cell types, the method significantly improves distributional similarity and predictive quality in functional genomics tasks. This work represents a notable advancement in applying advanced machine learning architectures to solve specific challenges in biological data analysis.
Retrieval-Augmented Generation for Predicting Cellular Responses to Gene Perturbation
arXiv:2603.07233v2 Announce Type: replace Abstract: Predicting transcriptional responses to genetic perturbations is fundamental to functional genomics and therapeutic discovery. Recent deep learning models have shown promise in single-cell perturbation response prediction, but they typically generate each response in isolation, without explicitly leveraging related perturbations. We introduce PT-RAG (Perturbation-aware Two-stage Retrieval-Augmented Generation), a plug-in retrieval-and-conditioning module for generative cellular perturbation response. PT-RAG augments an existing perturbation-response backbone with learned access to related perturbation contexts. The key challenge is that relevance is not fixed in this setting: functionally related genes may elicit different effects across cell types. PT-RAG addresses this with a two-stage retrieval mechanism: GenePT-based semantic retrieval first identifies K candidate perturbations, after which a differentiable Gumbel-Softmax selector adaptively selects retrieved contexts conditioned on the control cell state, the query perturbation, and each candidate perturbation. We evaluate PT-RAG on two backbones, a STATE-style generator used as a frozen random reservoir and a fully trained scGPT, across cross-cell-type and cross-perturbation generalization tasks. PT-RAG consistently improves distributional similarity and often overall predictive quality; for example, on scGPT cross-cell-type results, the 2-Wasserstein distance drops by 5.9% relative to scGPT alone. The code to reproduce our experiments is available at https://github.com/difra100/PT-RAG_NIPS.
Abstract:Predicting transcriptional responses to genetic perturbations is fundamental to functional genomics and therapeutic discovery. Recent deep learning models have shown promise in single-cell perturbation response prediction, but they typically generate each response in isolation, without explicitly leveraging related perturbations. We introduce PT-RAG (Perturbation-aware Two-stage Retrieval-Augmented Generation), a plug-in retrieval-and-conditioning module for generative cellular perturbation response. PT-RAG augments an existing perturbation-response backbone with learned access to related perturbation contexts. The key challenge is that relevance is not fixed in this setting: functionally related genes may elicit different effects across cell types. PT-RAG addresses this with a two-stage retrieval mechanism: GenePT-based semantic retrieval first identifies K candidate perturbations, after which a differentiable Gumbel-Softmax selector adaptively selects retrieved contexts conditioned on the control cell state, the query perturbation, and each candidate perturbation. We evaluate PT-RAG on two backbones, a STATE-style generator used as a frozen random reservoir and a fully trained scGPT, across cross-cell-type and cross-perturbation generalization tasks. PT-RAG consistently improves distributional similarity and often overall predictive quality; for example, on scGPT cross-cell-type results, the 2-Wasserstein distance drops by 5.9% relative to scGPT alone. The code to reproduce our experiments is available at this https URL.
Accepted to NeurIPS 2026 main track. 34 pages, 11 figures, 18 tables
Machine Learning (cs.LG); Information Retrieval (cs.IR)
arXiv:2603.07233 [cs.LG]
(or arXiv:2603.07233v2 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2603.07233 Focus to learn more arXiv-issued DOI via DataCite
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 300 words
Related coverage
CisTransCell: Single-Cell Perturbation Prediction via Gene Function, Regulatory Control, and Cellular Context
arXiv · Jun 15, 2026
From Cellular Responses to Pharmacological Domains: Multimodal Zero-Shot Drug Representation Learning
arXiv · Jul 29, 2026
Generalizable single-cell perturbation response prediction using energy-guided flow matching
arXiv · Oct 5, 2026
PerturbCellRL: Aligning Distributions and Grounding Biology via Post-Training Perturbation Generators
arXiv · Oct 5, 2026