Probing dynamical heterogeneities, fine structures, and missense mutations in Intrinsically Disordered Proteins using Molecular dynamics, MC simulations, and Deep Learning

1. We use a combination of Brownian dynamics (BD), Monte Carlo (MC) simulation, and Deep Learning (DL) strategies to study dynamical heterogeneities, fine structures, and missense mutations in intrinsically disordered proteins (IDPs). We use coarse-grained (CG) models where amino acids are represented as single beads interacting with hydropathy matrices and screened Coulomb interaction. First, we use two disparate hydropathy scales, HPS1 [1] and HPS2 [2], to conclude that for the optimized interaction strength, each model reproduces experimental gyration radii of the IDPs of length We investigate (i) universal aspects of the IDPs [3] to demonstrate that IDPs are broadly characterized with a Flory exponent of . We introduce (ii) Wilson charge index (W) and (ii) a skewness index (S) to capture additional properties in sequence and physical space. Finally, we study missense mutations implicated in Alzheimer's, Parkinson's, and other such diseases using a multi-layer perception neural net (NN) architecture with > accuracy and 10,000 -fold faster compared to BD simulation time, enabling us to study all possible missense mutations in a single IDP and identify potentially harmful mutations compared to wild types [5]. Our study will promote further research on more efficient and accurate predictions of lethal mutation effects in disordered proteins involved in Tauopathy, a-Synucleinopathy, etc. Finally, we compare BD simulation results for a two-bead IDP model (SOP-IDP) [7] with those from single-bead HPS models to assess the effect of the sidechains in promoting disorder in IDPs. We also elucidate the differences between synthetic and naturally occurring IDPs [5].

(*) Work done in collaboration with Swarnadeep Seth

1. 1. Dignon et al., PLOS Comp. Biology 14, e1005941 (2018).
   
   2. Tesei et al. PNAS 118, e2111696118 (2021).
   
   3. Jacob Bair, Swarnadeep Seth, and Aniket Bhattacharya, J. Chem. Phys. 158, 204902 (2023)
   
   4. Swarnadeep Seth and Aniket Bhattacharya, J. Chem. Phys. 160, 014902 (2024)
   
   5. Lotthamer, J.M., et al., Nature Methods, 21, 465 (2024); G. Tesei et al., Nature 626, 897 (2024).
   
   6. Swarnadeep Seth and Aniket Bhattacharya, bioRxiv 2024.07.07.602404
   
   7. U. Baul et al. J. Phys. Chem. B 123, 3462–3474 (2019).