Medium-sized protein language models perform well at transfer learning on realistic datasets
Luiz Carlos Vieira; Morgan L. Handojo; Claus O. Wilke · 2025 · Scientific Reports
WASTE classifies this as Negative / Null Result Report · AI classification, approximate
The study found no significant effect — useful as a negative control or null benchmark for your own design.
Abstract
Protein language models (pLMs) can offer deep insights into evolutionary and structural properties of proteins. While larger models, such as the 15 billion parameter model ESM-2, promise to capture more complex patterns in sequence space, they also present practical challenges due to their high dimensionality and high computational cost. We systematically evaluated the performance of various ESM-style models across multiple biological datasets to assess the impact of model size on transfer learning via feature extraction. Surprisingly, we found that larger models do not necessarily outperform
Abstract by Luiz Carlos Vieira; Morgan L. Handojo; Claus O. Wilke, Scientific Reports (2025) — licensed CC BY 4.0.
About to run something similar?
Run an AI Precheck on your own design to catch failure modes like this one before you spend the time. Your first desk check is free.
Related failures
Power and Predictive Accuracy of Polygenic Risk Scores
Negative / Null Result ReportLocoregional Recurrence After Sentinel Lymph Node Dissection With or Without Axillary Dissection in Patients With Sentinel Lymph Node Metastases
Negative / Null Result ReportCritical aspects of using bacterial cell viability assays with the fluorophores SYTO9 and propidium iodide
Negative / Null Result ReportAlpelisib plus fulvestrant for PIK3CA-mutated, hormone receptor-positive, human epidermal growth factor receptor-2–negative advanced breast cancer: final overall survival results from SOLAR-1
Negative / Null Result ReportA randomized placebo-controlled trial of idebenone in Leber’s hereditary optic neuropathy
Negative / Null Result ReportThe CAFA challenge reports improved protein function prediction and new functional annotations for hundreds of genes through experimental screens
WASTE indexes this work — it does not host or republish it. Failure-type classification is automated and approximate.
Metadata source: OpenAlex · DOI 10.1038/s41598-025-05674-x