CAMELOT: A machine learning approach for coarse-grained simulations of aggregation of block-copolymeric protein sequences

Abstract

We report the development and deployment of a coarse-graining method that is well suited for computer simulations of aggregation and phase separation of protein sequences with block-copolymeric architectures. Our algorithm, named CAMELOT for Coarse-grained simulations Aided by MachinE Learning Optimization and Training, leverages information from converged all atom simulations that is used to determine a suitable resolution and parameterize the coarse-grained model. To parameterize a system-specific coarse-grained model, we use a combination of Boltzmann inversion, non-linear regression, and a Gaussian process Bayesian optimization approach. The accuracy of the coarsegrained model is demonstrated through direct comparisons to results from all atom simulations. We demonstrate the utility of our coarse-graining approach using the block-copolymeric sequence from the exon 1 encoded sequence of the huntingtin protein. This sequence comprises of 17 residues from the N-terminal end of huntingtin (N17) followed by a polyglutamine (polyQ) tract. Simulations based on the CAMELOT approach are used to show that the adsorption and unfolding of the wild type N17 and its sequence variants on the surface of polyQ tracts engender a patchy colloid like architecture that promotes the formation of linear aggregates. These results provide a plausible explanation for experimental observations, which show that N17 accelerates the formation of linear aggregates in block-copolymeric N17-polyQ sequences. The CAMELOT approach is versatile and is generalizable for simulating the aggregation and phase behavior of a range of block-copolymeric protein sequences. C 2015 AIP Publishing LLC. [http://dx

LASSI: A lattice model for simulating phase transitions of multivalent proteins

“…These approaches allow for hybrid representations and can be generalized to systems of arbitrary complexity. What is required is the development of approaches that enable systematic coarsegraining and adaptation of machine learning based methods to parameterize interaction potentials (85). Engineering LASSI to be interoperable to cell-based modeling suites (86) will also allow for larger scale deployment of the overall framework.…”

Charge pattern matching as a ‘fuzzy’ mode of molecular recognition for the functional phase separations of intrinsically disordered proteins

“…Although computational study of IDPs is still in its infancy, much insight into the conformational properties and binding energetics of individual IDPs has been gained by explicit-chain simulations [8,9,[44][45][46][47][48][49]. In contrast, because IDP phase separation is a multiple-chain property, computationally it is extremely costly to simulate using fully atomic explicit-chain models [50], notwithstanding promising progress made by coarse-grained approaches that treat groups of amino acid residues of IDPs as interaction modules in continuum space [51] or on lattices [27] and simulation algorithms developed recently [52,53] for phase separations of globular proteins [54]. In this context, analytical theories of IDP phase separation are valuable not only because of their tractability, but also-and more importantly-for the conceptual framework they offer for understanding a highly complex phenomenon.…”

Phase Separation and Single-Chain Compactness of Charged Disordered Proteins Are Strongly Correlated

“…If we had adopted an unphysical interaction scheme without such a cutoff, similar trends would still hold although the scaling relations Eqs. ther explored by both theory and simulation (31,32) to help decipher the sequence determinants of IDP phase separation.…”