S-Pred: protein structural property prediction using MSA transformer

Abstract

Predicting the local structural features of a protein from its amino acid sequence helps its function prediction to be revealed and assists in three-dimensional structural modeling. As the sequence-structure gap increases, prediction methods have been developed to bridge this gap. Additionally, as the size of the structural database and computing power increase, the performance of these methods have also significantly improved. Herein, we present a powerful new tool called S-Pred, which can predict eight-state secondary structures (SS8), accessible surface areas (ASAs), and intrinsically disordered regions (IDRs) from a given sequence. For feature prediction, S-Pred uses multiple sequence alignment (MSA) of a query sequence as an input. The MSA input is converted to features by the MSA Transformer, which is a protein language model that uses an attention mechanism. A long short-term memory (LSTM) was employed to produce the final prediction. The performance of S-Pred was evaluated on several test sets, and the program consistently provided accurate predictions. The accuracy of the SS8 prediction was approximately 76%, and the Pearson’s correlation between the experimental and predicted ASAs was 0.84. Additionally, an IDR could be accurately predicted with an F1-score of 0.514. The program is freely available at https://github.com/arontier/S_Pred_Paper and https://ad3.io as a code and a web server.

Assessment of Disordered Linker Predictions in the CAID2 Experiment

“…CAID experiments provide invaluable insights into the performance of the current predictors of disorder and disorder functions [ 31 , 33 ]. They were recently used to identify accurate and practical methods for the prediction of disorder and disordered binding regions [ 62 ], analyze the performance of predictors that apply deep learning [ 58 ], investigate the use of AlphaFold for the disorder prediction [ 60 , 63 , 64 ], and design new methods for protein structure and disorder function predictions [ 47 , 65 , 66 , 67 , 68 , 69 , 70 , 71 , 72 , 73 , 74 ]. CAID2 featured the first evaluation of predictions of DLs [ 33 ].…”

Retrieved Sequence Augmentation for Protein Representation Learning

“…In protein learning, a similar approach Multiple Sequence Alignment (MSA) has been adopted to introduce evolutionary knowledge into models by augmenting input with aligned homologous sequences. MSA has improved deep learning performance on various models (Rao et al, 2021; Jumper et al, 2021; Marks et al, 2011; Hong et al, 2022), yet its success is often attributed to the alignment process that highlights co-evolution – especially the alignment process that is central to direct-coupling analysis methods (Morcos et al, 2011; Marks et al, 2011; Kamisetty et al, 2013). The most common practice for constructing MSA (Remmert et al, 2012; Altschul & Koonin, 1998; Johnson et al, 2010) is to build a Hidden Markov Model (HMM) profile for the entire sequence space of databases and then iteratively search for homologous sequences.…”