MI-PEFT: Mixture-of-Experts Integrated Parameter-Efficient Fine-Tuning Protein Language Models Improves Acidophilic Proteins Classification
Abstract
Acidophilic proteins that remain stable and functional under highly acidic conditions, are important for industrial biocatalysis, acid-related bioprocessing, and the discovery of acid-stable enzymes. However, their identification relies heavily on time-consuming experimental screening methods. With the rapid growth of protein sequence databases, the need for computational identification methods that are both accurate and efficient has become stronger. The emergence of protein language models (PLMs) has significantly improved the sequence representation of downstream biological prediction tasks. This paper proposes MI-PEFT, a mixture-of-experts integrated parameter-efficient fine-tuning framework. Built on the ESM C-600M backbone, the framework incorporates LoRA-based PEFT methods and a DeepSeekMoE-based classification head to resolve the limitations of PEFT and significantly improve computational efficiency. Notably, this task is characterized by a significant class imbalance in the dataset, making high specificity particularly challenging. The experimental results demonstrate that MI-PEFT on PLMs, especially {\text{C}}^{\text{3}}\text{A}, serves as an efficient tool for identifying acidophilic proteins and a constrained pathway that helps resolve class-imbalance by preserving the pretrained representations.
Related
- FedEpitope: Parameter-Efficient Federated Protein Language Models for Linear B-Cell Epitope Prediction
- Parameter-Efficient Adaptation of Protein Language Models for Dual-Target Antimicrobial Peptide Discovery Under Strict Homology Constraints
- EG-BINN: Parameter-Efficient Ensembling of Gated Biologically Informed Neural Networks for Epigenetic Age Estimation
- LiteMedCoT-VL: Parameter-Efficient Adaptation for Medical Visual Question Answering
- Democratizing Clinical Tumor Whole Genome Sequencing: 18-hour End-to-end Analysis via Trillion-parameter Large Language Models Locally Deployed on Consumer-grade Hardware
BibTeX
@misc{shen2026mipeft,
title = {{MI-PEFT: Mixture-of-Experts Integrated Parameter-Efficient Fine-Tuning Protein Language Models Improves Acidophilic Proteins Classification}},
author = {Honghan Shen},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.08059},
eprint = {2609.08059},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.08059},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - MI-PEFT: Mixture-of-Experts Integrated Parameter-Efficient Fine-Tuning Protein Language Models Improves Acidophilic Proteins Classification
AU - Honghan Shen
PY - 2026
DA - 2026/09/07
PB - arXiv
DO - 10.48550/arxiv.2609.08059
UR - https://arxiv.org/abs/2609.08059
AB - Acidophilic proteins that remain stable and functional under highly acidic conditions, are important for industrial biocatalysis, acid-related bioprocessing, and the discovery of acid-stable enzymes. However, their identification relies heavily on time-consuming experimental screening methods. With the rapid growth of protein sequence databases, the need for computational identification methods that are both accurate and efficient has become stronger. The emergence of protein language models (PLMs) has significantly improved the sequence representation of downstream biological prediction tasks. This paper proposes MI-PEFT, a mixture-of-experts integrated parameter-efficient fine-tuning framework. Built on the ESM C-600M backbone, the framework incorporates LoRA-based PEFT methods and a DeepSeekMoE-based classification head to resolve the limitations of PEFT and significantly improve computational efficiency. Notably, this task is characterized by a significant class imbalance in the dataset, making high specificity particularly challenging. The experimental results demonstrate that MI-PEFT on PLMs, especially {\text{C}}^{\text{3}}\text{A}, serves as an efficient tool for identifying acidophilic proteins and a constrained pathway that helps resolve class-imbalance by preserving the pretrained representations.
ER -