Preprint · not peer-reviewed Researchers Educators Students

Preprint introduces LAMP framework to improve how genomic sequence-to-function models learn variant effects

A preprint describes LAMP, a post-training approach that addresses systematic underestimation of noncoding variant effects in sequence-to-function models used for gene expression prediction.

Published · AI-drafted summary based on 1 public source
Illustration for polygenic story
Illustrative image — not from the source article.
Share

A preprint posted to bioRxiv describes LAMP, a framework designed to improve genomic sequence-to-function (S2F) models during post-training. S2F models predict molecular phenotypes — such as chromatin accessibility or gene expression — from DNA sequence, and are increasingly used to interpret noncoding genetic variation in the context of genome-wide association studies and functional genomics.

The authors identify a specific limitation: current S2F models systematically underestimate the effects of most causal noncoding variants, which causes fine-mapping failures and means these models underperform relative to simpler linear baselines when predicting gene expression from personal genome sequences. LAMP is presented as a post-training intervention that improves personal genome predictions by addressing this underestimation, offering better calibration without retraining models from scratch.

The work is relevant to researchers using computational tools to interpret noncoding variants, particularly in GWAS follow-up and functional prioritisation pipelines. As a preprint on bioRxiv, the findings have not yet undergone peer review. The authors' framing situates LAMP within the ongoing challenge of moving from association to function in complex trait genetics — a problem that remains active across the field.

Sources

Read the original reporting — these are the public sources this summary draws from.

  1. Primary sourcePreprint bioRxiv (Cold Spring Harbor Laboratory) · 2026-11-17
    LAMP sheds light on how sequence to function models learn variant effects during post-training

Tags

sequence-to-function-models noncoding-variants gene-expression gwas functional-genomics post-gwas computational-genomics preprint
Share

About Genetic Current

Educational summaries of public genetics news

Genetic Current is the news section of Evagene, an academic, research, and educational pedigree-modelling platform. Stories are AI-drafted summaries of items from trusted public sources, written for researchers, clinicians, educators, students, genealogists, and patients with an interest in genetics. Summaries are for educational and research purposes only and are not medical advice.

Join the Evagene Alpha Waiting List