Precise Generation of Conformational Ensembles for Intrinsically Disordered Proteins via Fine-tuned Diffusion Models

Abstract

Intrinsically disordered proteins (IDPs) play pivotal roles in various biological functions and are closely linked to many human diseases including cancer. Structural investigations of IDPs typically involve a combination of molecular dynamics (MD) simulations and experimental data to correct for intrinsic biases in simulation methods. However, these simulations are hindered by their high computational cost and a scarcity of experimental data, severely limiting their applicability. Despite the recent advancements in structure prediction for structured proteins, understanding the conformational properties of IDPs remains challenging partly due to the poor conservation of disordered protein sequences and limited experimental characterization. Here, we introduced IDPFold, a method capable of predicting IDP conformation ensembles directly from their sequences using fine-tuned diffusion models. IDPFold bypasses the need for Multiple Sequence Alignments (MSA) or experimental data, achieving accurate predictions of ensemble properties across numerous IDPs. By sampling conformations at the backbone level, IDPFold provides more detailed structural features and more precise property estimation compared to the state-of-the-art methods, and will help to reveal the disorder-function paradigm of IDPs.

Sampling Conformational Ensembles of Highly Dynamic Proteins via Generative Deep Learning

“…Artificial intelligence (AI) provides an alternative approach for accelerating the generation of protein conformational ensembles. Novel machine learning/deep learning (ML/DL) approaches analyze simulation results to further guide conformational search with enhanced sampling techniques [19][20][21][22][23][24][25][26][27][28] . ML/DL optimizes coarse-grain models to speed up conformation transitions with preserved atomistic interactions 29,30 .…”