Sequence-sensitive elastic network captures dynamical features necessary for miR-125a maturation

Abstract

The Elastic Network Contact Model (ENCoM) is a coarse-grained normal mode analysis (NMA) model unique in its all-atom sensitivity to the sequence of the studied macromolecule and thus to the effect of mutations. We adapted ENCoM to simulate the dynamics of ribonucleic acid (RNA) molecules, benchmarked its performance against other popular NMA models and used it to study the 3D structural dynamics of human microRNA miR-125a, leveraging high-throughput experimental maturation efficiency data of over 26 000 sequence variants. We also introduce a novel way of using dynamical information from NMA to train multivariate linear regression models, with the purpose of highlighting the most salient contributions of dynamics to function. ENCoM has a similar performance profile on RNA than on proteins when compared to the Anisotropic Network Model (ANM), the most widely used coarse-grained NMA model; it has the advantage on predicting large-scale motions while ANM performs better on B-factors prediction. A stringent benchmark from the miR-125a maturation dataset, in which the training set contains no sequence information in common with the testing set, reveals that ENCoM is the only tested model able to capture signal beyond the sequence. This ability translates to better predictive power on a second benchmark in which sequence features are shared between the train and test sets. When training the linear regression model using all available data, the dynamical features identified as necessary for miR-125a maturation point to known patterns but also offer new insights into the biogenesis of microRNAs. Our novel approach combining NMA with multivariate linear regression is generalizable to any macromolecule for which relatively high-throughput mutational data is available.

The DynaSig-ML Python package: automated learning of biomolecular dynamics–function relationships

“…microRNAs are short single-stranded RNAs of ∼22 nucleotides which regulated gene expression by guiding the RNA-induced silencing complex to complementary regions within messenger RNAs. In our recent work, we adapted ENCoM to work on RNA molecules and used it to study dynamics–function relationships apparent from an experimental mutagenesis dataset ( Fang and Bartel 2015 ) of over 29 000 sequence variants of miR-125a, a human microRNA ( Mailhot et al 2022 ). In order to illustrate a typical use case of DynaSig-ML, we applied it to study dynamics–function relationships in miR-125a sequence variants, replicating the results from our work in an automated way.…”

“…The results reported in Fig. 1 use our inverted dataset previously describes ( Mailhot et al 2022 ), in which the training set contains variants with only one or two mutations and the testing set contains variants with three to six mutations. It tests the models’ ability to generalize to variants containing more mutations than what was seen in training, which is very relevant in the context of using DynaSig-ML for high-throughput in silico predictions.…”

“…Its sequence sensitivity enables its use to predict the impact of sequence variants on biomolecular function through changes in predicted stability ( Frappier and Najmanovich 2015 ) and dynamics ( Teruel et al 2021 ). We recently extended ENCoM to work on RNA molecules and predicted microRNA maturation efficiency from a dataset of experimentally measured maturation efficiencies of over 26 000 sequence variants using LASSO regression ( Mailhot et al 2022 ). To do so, the ENCoM Dynamical Signatures, which are vectors of predicted structural fluctuations at every position in the system, were used as input variables in a LASSO multiple linear regression model ( Tibshirani 1996 ) to predict maturation efficiency.…”

“…Considering that ENCoM can be used to study proteins, nucleic acids, small molecules, and their complexes ( Mailhot and Najmanovich 2021 ), DynaSig-ML can be applied to any biomolecule for which there exist experimental data linking perturbations (such as mutations or ligand binding) to function. To demonstrate the use of DynaSig-ML, we predict maturation efficiencies of miR-125a sequence variants ( Mailhot et al 2022 ), exploring gradient boosting, and random forest (RF) regressors in addition to LASSO regression. An accompanying online step-by-step tutorial takes users through the necessary steps to generate all results shown in the present work.…”

“…In the case of LASSO regression or other forms of regression, the learned coefficients can be automatically mapped back on the studied structure by DynaSig-ML and visualized in 3D with two simple PyMOL ( Delano 2002 ) commands. These coefficients represent the relationship between flexibility changes at specific positions and the predicted experimental property so the mapping can be used to drive new biological hypotheses ( Mailhot et al 2022 ). DynaSig-ML also automatically generates graphs showing the performance of each machine learning algorithm test.…”