2025 Hyper Recent •CC0 1.0 Universal

This work is dedicated to the public domain. No rights reserved.

Access Preprint From Server
May 8th, 2025
Version: 1
Duke University
bioengineering
biorxiv

Functional alignment of protein language models via reinforcement learning

Blalock, N.Open in Google Scholar•Seshadri, S.Open in Google Scholar•Babbar, A.Open in Google Scholar•Fahlberg, S. A.Open in Google Scholar•Kulkarni, A.Open in Google Scholar•Romero, P. A.Open in Google Scholar

Protein language models (pLMs) enable generative design of novel protein sequences but remain fundamentally misaligned with protein engineering goals, as they lack explicit understanding of function and often fail to improve properties beyond those found in nature. We introduce Reinforcement Learning from eXperimental Feedback (RLXF), a general framework that aligns protein language models with experimentally measured functional objectives, drawing inspiration from the methods used to align large language models like ChatGPT. Applied across five diverse protein families, RLXF improves generation of high-functioning variants beyond pre-trained baselines. We demonstrate this with CreiLOV, an oxygen-independent fluorescent protein, where RLXF-aligned models generate sequences with significantly enhanced fluorescence, including the most fluorescent CreiLOV variants reported to date. Our results indicate that RLXF-aligned models effectively integrate the evolutionary knowledge encoded in pre-trained pLMs with experimental observations, improving the success rate of generated sequences and enabling the discovery of synergistic mutation combinations that are difficult to identify through zero-shot or evolutionary approaches. RLXF provides a scalable and accessible approach to steer generative models toward desired biochemical properties, enabling function-driven protein design beyond the limits of natural evolution.

Similar Papers

biorxiv
Fri May 09 2025
A plug-and-play transepithelial/transendothelial electric resistance (TEER)-upgraded organ-on-chip system to measure barrier dynamics in real-time
The integration of transepithelial/transendothelial electrical resistance (TEER) measurement into organ-on-chip (OoC) platforms provides a unique opportunity to monitor the integrity of biological barriers in real-time. This is particularly important for detecting rapid changes in the temporal dynamics of intercellular junctional complexes in response to drug compounds, changes in host-microbiota ...
Kaden, T.
•
Besser, S.
•
Abdo, N.
•
Mosig, A. S.
...•
Nietzsche, S.
biorxiv
Fri May 09 2025
Erasable Synthetic Serum Markers
Gene expression in the brain is typically evaluated using invasive biopsy or postmortem histology. Serum markers provide an alternative way to monitor the brain, but relatively few such markers exist. Additionally, the origin of serum markers often cannot be localized to a specific cell population, and monitoring dynamic changes in their gene expression is compromised by the same factor that makes...
Nouraein, S.
•
Lee, S.
•
Li, H.
•
Saenz, V.
...•
Szablowski, J. O.
biorxiv
Thu May 08 2025
Harmonization of Structural Brain Connectivity through Distribution Matching
The increasing prevalence of multi-site diffusion-weighted magnetic resonance imaging (dMRI) studies potentially offers enhanced statistical power to investigate brain structure. However, these studies face challenges due to variations in scanner hardware and acquisition protocols. While several methods for dMRI data harmonization exist, few specifically address structural brain connectivity. We i...
Zhou, Z.
•
Fischl, B.
•
Aganj, I.
biorxiv
Thu May 08 2025
NeuroMark-HiFi: A Data-Driven Method for Detecting High-Spatial-Frequency Functional Brain Networks
Objective: The Traditional functional neuroimaging approaches typically focus on low-frequency spatial structures, potentially overlooking critical fine-scale connectivity disruptions associated with brain disorders. Methods: We introduce NeuroMark-HiFi, a fully automated algorithm designed to enhance the detection of high-spatial-frequency functional brain network patterns. NeuroMark-HiFi systema...
Behzadfar, N.
•
Iraji, A.
•
Calhoun, V.
biorxiv
Wed May 07 2025
Fetal Health Classification Based on CTG
This research paper explores the application of advanced machine learning techniques for fetal health detection using cardiotocography (CTG). Cardiotocography is a pivotal tool in monitoring fetal and maternal health during pregnancy, providing crucial insights into fetal heart rate patterns and uterine contractions. In this work, various predictive models, including logistic regression, nearest n...
Madiraju, R.
•
Upadhyay, U.
•
C, M.
biorxiv
Wed May 07 2025
Multimodal imaging reveals multiscale mechanical interplay in vertebral endplate microarchitecture during intervertebral disc loading
The function of all musculoskeletal joints depends on hierarchical structures spanning the molecular to whole joint scales. Investigating biomechanics across length scales requires correlative multiscale experimental methods. This study applies multimodal in situ synchrotron imaging techniques to spinal joints, focussing on the vertebral endplates, to explore relationships between structure and me...
Parmenter, A. L.
•
Newham, E.
•
Sharma, A.
•
Disney, C. M.
...•
Lee, P. D.
biorxiv
Wed May 07 2025
3D Printed Nerve Guidance Conduit for Biologics-Free Nerve Regeneration and Vascular Integration
There is a clinical need for an effective nerve guidance conduit to treat peripheral nerve injuries. Many studies have explored different materials and active cues to guide neural regeneration, with some success. However, none have demonstrated a comparable or better functional recovery than the clinical standard autograft. Autografts are often insufficient for reconstruction of an injury to long ...
Schimelman, J.
•
Berry, D. B.
•
Johnson, S.
•
Shi, R.
...•
Chen, S.
biorxiv
Wed May 07 2025
Combined optical coherence tomography and electroretinography (OCT+ERG) system for imaging neurovascular coupling in the human retina
Significance: During their early stages of development, neurological and neurodegenerative diseases cause changes to the biological tissue's morphology, physiology and metabolism at cellular level, and acute, transient changes in the local blood flow. Development of novel optical methods for quantitative imaging of such changes non-invasively and simultaneously would allow for probing of neurovasc...
Dhaliwal, K. K.
•
Wong, A.
•
Wright, T.
•
Bizheva, K.
biorxiv
Wed May 07 2025
Closed-loop sonothermogenetic control of CAR T cells for metronomic brain cancer therapy
Achieving durable CAR T cell responses against primary brain tumors and metastases requires strategies that enable intracranial control of therapy to overcome the barriers of solid tumor treatment without compromising safety. Here, we show that closed-loop sonothermogenetics enables remote regulation of CAR T cell therapeutic activity through the intact skull. Using MR-guided focused ultrasound wi...
Zamat, A.
•
Kim, C.
•
Sridhar, S.
•
Fabrega, S.
...•
Kwong, G. A.