TMbed: transmembrane proteins predicted through language model embeddings

BACKGROUND

Despite the immense importance of transmembrane proteins (TMP) for molecular biology and medicine, experimental 3D structures for TMPs remain about 4-5 times underrepresented compared to non-TMPs. Today's top methods such as AlphaFold2 accurately predict 3D structures for many TMPs, but annotating transmembrane regions remains a limiting step for proteome-wide predictions.

RESULTS

Here, we present TMbed, a novel method inputting embeddings from protein Language Models (pLMs, here ProtT5), to predict for each residue one of four classes: transmembrane helix (TMH), transmembrane strand (TMB), signal peptide, or other. TMbed completes predictions for entire proteomes within hours on a single consumer-grade desktop machine at performance levels similar or better than methods, which are using evolutionary information from multiple sequence alignments (MSAs) of protein families. On the per-protein level, TMbed correctly identified 94 ± 8% of the beta barrel TMPs (53 of 57) and 98 ± 1% of the alpha helical TMPs (557 of 571) in a non-redundant data set, at false positive rates well below 1% (erred on 30 of 5654 non-membrane proteins). On the per-segment level, TMbed correctly placed, on average, 9 of 10 transmembrane segments within five residues of the experimental observation. Our method can handle sequences of up to 4200 residues on standard graphics cards used in desktop PCs (e.g., NVIDIA GeForce RTX 3060).

CONCLUSIONS

Based on embeddings from pLMs and two novel filters (Gaussian and Viterbi), TMbed predicts alpha helical and beta barrel TMPs at least as accurately as any other method but at lower false positive rates. Given the few false positives and its outstanding speed, TMbed might be ideal to sieve through millions of 3D structures soon to be predicted, e.g., by AlphaFold2.

Hidden neural networks for transmembrane protein topology prediction

Hidden Markov Models (HMMs) are amongst the most successful methods for predicting protein features in biological sequence analysis. However, there are biological problems where the Markovian assumption is not sufficient since the sequence context can provide useful information for prediction purposes. Several extensions of HMMs have appeared in the literature in order to overcome their limitations. We apply here a hybrid method that combines HMMs and Neural Networks (NNs), termed Hidden Neural Networks (HNNs), for biological sequence analysis in a straightforward manner. In this framework, the traditional HMM probability parameters are replaced by NN outputs. As a case study, we focus on the topology prediction of for alpha-helical and beta-barrel membrane proteins. The HNNs show performance gains compared to standard HMMs and the respective predictors outperform the top-scoring methods in the field. The implementation of HNNs can be found in the package JUCHMME, downloadable from http://www.compgen.org/tools/juchmme, https://github.com/pbagos/juchmme. The updated PRED-TMBB2 and HMM-TM prediction servers can be accessed at www.compgen.org.

JUCHMME: a Java Utility for Class Hidden Markov Models and Extensions for biological sequence analysis

SUMMARY

JUCHMME is an open-source software package designed to fit arbitrary custom Hidden Markov Models (HMMs) with a discrete alphabet of symbols. We incorporate a large collection of standard algorithms for HMMs as well as a number of extensions and evaluate the software on various biological problems. Importantly, the JUCHMME toolkit includes several additional features that allow for easy building and evaluation of custom HMMs, which could be a useful resource for the research community.

AVAILABILITY AND IMPLEMENTATION

http://www.compgen.org/tools/juchmme, https://github.com/pbagos/juchmme.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Semi-supervised learning of Hidden Markov Models for biological sequence analysis

MOTIVATION

Hidden Markov Models (HMMs) are probabilistic models widely used in applications in computational sequence analysis. HMMs are basically unsupervised models. However, in the most important applications, they are trained in a supervised manner. Training examples accompanied by labels corresponding to different classes are given as input and the set of parameters that maximize the joint probability of sequences and labels is estimated. A main problem with this approach is that, in the majority of the cases, labels are hard to find and thus the amount of training data is limited. On the other hand, there are plenty of unclassified (unlabeled) sequences deposited in the public databases that could potentially contribute to the training procedure. This approach is called semi-supervised learning and could be very helpful in many applications.

RESULTS

We propose here, a method for semi-supervised learning of HMMs that can incorporate labeled, unlabeled and partially labeled data in a straightforward manner. The algorithm is based on a variant of the Expectation-Maximization (EM) algorithm, where the missing labels of the unlabeled or partially labeled data are considered as the missing data. We apply the algorithm to several biological problems, namely, for the prediction of transmembrane protein topology for alpha-helical and beta-barrel membrane proteins and for the prediction of archaeal signal peptides. The results are very promising, since the algorithms presented here can significantly improve the prediction performance of even the top-scoring classifiers.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.