Multi-location gram-positive and gram-negative bacterial protein subcellular localization using gene ontology and multi-label classifier ensemble

A Review for Artificial Intelligence Based Protein Subcellular Localization

Proteins need to be located in appropriate spatiotemporal contexts to carry out their diverse biological functions. Mislocalized proteins may lead to a broad range of diseases, such as cancer and Alzheimer's disease. Knowing where a target protein resides within a cell will give insights into tailored drug design for a disease. As the gold validation standard, the conventional wet lab uses fluorescent microscopy imaging, immunoelectron microscopy, and fluorescent biomarker tags for protein subcellular location identification. However, the booming era of proteomics and high-throughput sequencing generates tons of newly discovered proteins, making protein subcellular localization by wet-lab experiments a mission impossible. To tackle this concern, in the past decades, artificial intelligence (AI) and machine learning (ML), especially deep learning methods, have made significant progress in this research area. In this article, we review the latest advances in AI-based method development in three typical types of approaches, including sequence-based, knowledge-based, and image-based methods. We also elaborately discuss existing challenges and future directions in AI-based method development in this research field.

ML-FGAT: Identification of multi-label protein subcellular localization by interpretable graph attention networks and feature-generative adversarial networks

The prediction of multi-label protein subcellular localization (SCL) is a pivotal area in bioinformatics research. Recent advancements in protein structure research have facilitated the application of graph neural networks. This paper introduces a novel approach termed ML-FGAT. The approach begins by extracting node information of proteins from sequence data, physical-chemical properties, evolutionary insights, and structural details. Subsequently, various evolutionary techniques are integrated to consolidate multi-view information. A linear discriminant analysis framework, grounded on entropy weight, is then employed to reduce the dimensionality of the merged features. To enhance the robustness of the model, the training dataset is augmented using feature-generative adversarial networks. For the primary prediction step, graph attention networks are employed to determine multi-label protein SCL, leveraging both node and neighboring information. The interpretability is enhanced by analyzing the attention weight parameters. The training is based on the Gram-positive bacteria dataset, while validation employs newly constructed datasets: human, virus, Gram-negative bacteria, plant, and SARS-CoV-2. Following a leave-one-out cross-validation procedure, ML-FGAT demonstrates noteworthy superiority in this domain.

Identification of risk profiles for Salmonella prevalence in pig supply chains in South Korea using meta-analysis and a quantitative microbial risk assessment model

International travel and the globalization of food supplies have increased the risk of epidemic foodborne infections. Salmonella strains, particularly non-typhoidal Salmonella (NTS), are major zoonotic pathogens responsible for gastrointestinal diseases worldwide. In this study, the prevalence and Salmonella contamination in pigs/carcasses throughout the South Korean pig supply chain and the associated risk factors were evaluated using Systematic reviews and meta-analyses (SRMA), and quantitative microbial risk assessment (QMRA). The prevalence of Salmonella in finishing pigs, which is one of the major starting inputs of the QMRA model was calculated through SRMA of studies conducted in south Korea in order to complement and enhance the robustness of the model. Our findings revealed that the pooled Salmonella prevalence in pigs was 4.15% with a 95% confidence interval (CI) of 2.56 to 6.66%. Considering the pig supply chain, the highest prevalence was detected in slaughterhouses (6.27% [95% CI: 3.36; 11.37]), followed by farms (4.16% [95% CI: 2.32; 7.35]) and meat stores (1.21% [95% CI: 0.42; 3.46]). The QMRA model predicted a 3.9% likelihood of Salmonella-free carcasses and a 96.1% probability of Salmonella-positive carcasses at the end of slaughter, with an average Salmonella concentration of 6.38 log CFU/carcass (95% CI: 5.17; 7.28). This corresponds to an average contamination of 1.23 log CFU/g (95% CI: 0.37; 2.48) of pork meat. Across the pig supply chain, the highest Salmonella contamination was predicted after transport and lairage, with an average concentration of 8 log CFU/pig (95% CI: 7.15; 8.42). Sensitivity analysis indicated that Salmonella fecal shedding (r = 0.68) and Salmonella prevalence in finishing pigs (r = 0.39) at pre-harvest were the most significant factors associated with Salmonella contamination in pork carcasses. Although disinfection and sanitation interventions along the slaughter line can reduce contamination levels to some extent, effective measures should be taken to reduce Salmonella prevalence at the farm level to improve the safety of pork consumption.

A robust protein language model for SARS-CoV-2 protein-protein interaction network prediction

Protein-protein interaction is one of the ways viruses interact with their hosts. Therefore, identifying protein interactions between viruses and hosts helps explain how virus proteins work, how they replicate, and how they cause disease. SARS-CoV-2 is a new type of virus that emerged from the coronavirus family in 2019 and caused a worldwide pandemic. Detection of human proteins interacting with this novel virus strain plays an important role in monitoring the cellular process of virus-associated infection. Within the scope of the study, a natural language processing-based collective learning method is proposed for the prediction of potential SARS-CoV-2-human PPIs. Protein language models were obtained with the prediction-based word2Vec and doc2Vec embedding methods and the frequency-based tf-idf method. Known interactions were represented by proposed language models and traditional feature extraction methods (conjoint triad and repeat pattern), and their performances were compared. The interaction data were trained with support vector machine, artificial neural network (ANN), k-nearest neighbor (KNN), naive Bayes (NB), decision tree (DT), and ensemble algorithms. Experimental results show that protein language models are a promising protein representation method for protein-protein interaction prediction. The term frequency-inverse document frequency-based language model performed the SARS-CoV-2 protein-protein interaction estimation with an error of 1.4%. Additionally, the decisions of high-performing learning models for different feature extraction methods were combined with a collective voting approach to make new interaction predictions. For 10,000 human proteins, 285 new potential interactions were predicted, with models combining decisions.

Diagnosis of autism spectrum disorder based on functional brain networks and machine learning

Autism is a multifaceted neurodevelopmental condition whose accurate diagnosis may be challenging because the associated symptoms and severity vary considerably. The wrong diagnosis can affect families and the educational system, raising the risk of depression, eating disorders, and self-harm. Recently, many works have proposed new methods for the diagnosis of autism based on machine learning and brain data. However, these works focus on only one pairwise statistical metric, ignoring the brain network organization. In this paper, we propose a method for the automatic diagnosis of autism based on functional brain imaging data recorded from 500 subjects, where 242 present autism spectrum disorder considering the regions of interest throughout Bootstrap Analysis of Stable Cluster map. Our method can distinguish the control group from autism spectrum disorder patients with high accuracy. Indeed the best performance provides an AUC near 1.0, which is higher than that found in the literature. We verify that the left ventral posterior cingulate cortex region is less connected to an area in the cerebellum of patients with this neurodevelopment disorder, which agrees with previous studies. The functional brain networks of autism spectrum disorder patients show more segregation, less distribution of information across the network, and less connectivity compared to the control cases. Our workflow provides medical interpretability and can be used on other fMRI and EEG data, including small data sets.

Pangenome-based genome inference allows efficient and accurate genotyping across a wide spectrum of variant classes

Typical genotyping workflows map reads to a reference genome before identifying genetic variants. Generating such alignments introduces reference biases and comes with substantial computational burden. Furthermore, short-read lengths limit the ability to characterize repetitive genomic regions, which are particularly challenging for fast k-mer-based genotypers. In the present study, we propose a new algorithm, PanGenie, that leverages a haplotype-resolved pangenome reference together with k-mer counts from short-read sequencing data to genotype a wide spectrum of genetic variation-a process we refer to as genome inference. Compared with mapping-based approaches, PanGenie is more than 4 times faster at 30-fold coverage and achieves better genotype concordances for almost all variant types and coverages tested. Improvements are especially pronounced for large insertions (≥50 bp) and variants in repetitive regions, enabling the inclusion of these classes of variants in genome-wide association studies. PanGenie efficiently leverages the increasing amount of haplotype-resolved assemblies to unravel the functional impact of previously inaccessible variants while being faster compared with alignment-based workflows.

An empirical test of the relationship between the bootstrap and likelihood ratio support in maximum likelihood phylogenetic analysis

In maximum likelihood (ML), the support for a clade can be calculated directly as the likelihood ratio (LR) or log-likelihood difference (S, LLD) of the best trees with and without the clade of interest. However, bootstrap (BS) clade frequencies are more pervasive in ML phylogenetics and are almost universally interpreted as measuring support. In addition to theoretical arguments against that interpretation, BS has several undesirable attributes for a support measure. For example, it does not vary in proportion to optimality or identify clades that are rejected by the evidence and can be overestimated due to missing data. Nevertheless, if BS is a reliable predictor of S, then it might be an efficient indirect method of measuring support-an attractive possibility, given the speed of many BS implementations. To assess the relationship between S and BS, we analyzed 106 empirical datasets retrieved from TreeBASE. Also, to evaluate the degree to which S and BS are affected by the number of replicates during suboptimal tree searches for S and pseudoreplicates during BS estimation, we randomly selected 5 of the 106 datasets and analyzed them using variable numbers of replicates and pseudoreplicates, respectively. The correlation between S and BS was extremely weak in the datasets we analyzed. Increasing the number of replicates during tree search decreased the estimated values of S for most clades, but the magnitude of change was small. In contrast, although increasing pseudoreplicates affected BS values for only approximately 40% of clades, values both increased and decreased, and they did so at much greater magnitudes. Increasing replicates/pseudoreplicates affected the rank order of clades in each tree for both S and BS. Our findings show decisively that BS is not an efficient indirect method of measuring support and suggest that even quite superficial searches to calculate S provide better estimates of support.

Predicting the multi-label protein subcellular localization through multi-information fusion and MLSI dimensionality reduction based on MLFE classifier

MOTIVATION

Multi-label (ML) protein subcellular localization (SCL) is an indispensable way to study protein function. It can locate a certain protein (such as the human transmembrane protein that promotes the invasion of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)) or expression product at a specific location in a cell, which can provide a reference for clinical treatment of diseases such as coronavirus disease 2019 (COVID-19).

RESULTS

The article proposes a novel method named ML-locMLFE. First of all, six feature extraction methods are adopted to obtain protein effective information. These methods include pseudo amino acid composition, encoding based on grouped weight, gene ontology, multi-scale continuous and discontinuous, residue probing transformation and evolutionary distance transformation. In the next part, we utilize the ML information latent semantic index method to avoid the interference of redundant information. In the end, ML learning with feature-induced labeling information enrichment is adopted to predict the ML protein SCL. The Gram-positive bacteria dataset is chosen as a training set, while the Gram-negative bacteria dataset, virus dataset, newPlant dataset and SARS-CoV-2 dataset as the test sets. The overall actual accuracy of the first four datasets are 99.23%, 93.82%, 93.24% and 96.72% by the leave-one-out cross validation. It is worth mentioning that the overall actual accuracy prediction result of our predictor on the SARS-CoV-2 dataset is 72.73%. The results indicate that the ML-locMLFE method has obvious advantages in predicting the SCL of ML protein, which provides new ideas for further research on the SCL of ML protein.

AVAILABILITY AND IMPLEMENTATION

The source codes and datasets are publicly available at https://github.com/QUST-AIBBDRC/ML-locMLFE/.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Three Distinct Proteases Are Responsible for Overall Cell Surface Proteolysis in Streptococcus thermophilus

The lactic acid bacterium Streptococcus thermophilus was believed to display only two distinct proteases at the cell surface, namely, the cell envelope protease PrtS and the housekeeping protease HtrA. Using peptidomics, we demonstrate here the existence of an additional active cell surface protease, which shares significant homology with the SepM protease of Streptococcus mutans. Although all three proteases-PrtS, HtrA, and SepM-are involved in the turnover of surface proteins, they demonstrate distinct substrate specificities. In particular, SepM cleaves proteins involved in cell wall metabolism and cell elongation, and its inactivation has consequences for cell morphology. When all three proteases are inactivated, the residual cell-surface proteolysis of S. thermophilus is approximately 5% of that of the wild-type strain. IMPORTANCE Streptococcus thermophilus is a lactic acid bacterium used widely as a starter in the dairy industry. Due to its "generally recognized as safe" status and its weak cell surface proteolytic activity, it is also considered a potential bacterial vector for heterologous protein production. Our identification of a new cell surface protease made it possible to construct a mutant strain with a 95% reduction in surface proteolysis, which could be useful in numerous biotechnological applications.

Accurate prediction of multi-label protein subcellular localization through multi-view feature learning with RBRL classifier

Multi-label proteins can participate in carrier transportation, enzyme catalysis, hormone regulation and other life activities. Meanwhile, they play a key role in the fields of biopharmaceuticals, gene and cell therapy. This article proposes a prediction method called Mps-mvRBRL to predict the subcellular localization (SCL) of multi-label protein. Firstly, pseudo position-specific scoring matrix, dipeptide composition, position specific scoring matrix-transition probability composition, gene ontology and pseudo amino acid composition algorithms are used to obtain numerical information from different views. Based on the contribution of five individual feature extraction methods, differential evolution is used for the first time to learn the weight of single feature, and then these original features use a weighted combination method to fuse multi-view information. Secondly, the fused high-dimensional features use a weighted linear discriminant analysis framework based on binary weight form to eliminate irrelevant information. Finally, the best feature vector is input into the joint ranking support vector machine and binary relevance with robust low-rank learning classifier to predict the SCL. After applying leave-one-out cross-validation, the overall actual accuracy (OAA) and overall location accuracy (OLA) of Mps-mvRBRL on the training set of Gram-positive bacteria are both 99.81%. The OAA on the test sets of plant, virus and Gram-negative bacteria datasets are 97.24%, 98.55% and 98.20%, respectively, and the OLA are 97.16%, 97.62% and 98.28%, respectively. The results show that the model achieves good prediction performance for predicting the SCL of multi-label protein.

SubLocEP: a novel ensemble predictor of subcellular localization of eukaryotic mRNA based on machine learning

MOTIVATION

mRNA location corresponds to the location of protein translation and contributes to precise spatial and temporal management of the protein function. However, current assignment of subcellular localization of eukaryotic mRNA reveals important limitations: (1) turning multiple classifications into multiple dichotomies makes the training process tedious; (2) the majority of the models trained by classical algorithm are based on the extraction of single sequence information; (3) the existing state-of-the-art models have not reached an ideal level in terms of prediction and generalization ability. To achieve better assignment of subcellular localization of eukaryotic mRNA, a better and more comprehensive model must be developed.

RESULTS

In this paper, SubLocEP is proposed as a two-layer integrated prediction model for accurate prediction of the location of sequence samples. Unlike the existing models based on limited features, SubLocEP comprehensively considers additional feature attributes and is combined with LightGBM to generated single feature classifiers. The initial integration model (single-layer model) is generated according to the categories of a feature. Subsequently, two single-layer integration models are weighted (sequence-based: physicochemical properties = 3:2) to produce the final two-layer model. The performance of SubLocEP on independent datasets is sufficient to indicate that SubLocEP is an accurate and stable prediction model with strong generalization ability. Additionally, an online tool has been developed that contains experimental data and can maximize the user convenience for estimation of subcellular localization of eukaryotic mRNA.

Tools for the Recognition of Sorting Signals and the Prediction of Subcellular Localization of Proteins From Their Amino Acid Sequences

At the time of translation, nascent proteins are thought to be sorted into their final subcellular localization sites, based on the part of their amino acid sequences (i.e., sorting or targeting signals). Thus, it is interesting to computationally recognize these signals from the amino acid sequences of any given proteins and to predict their final subcellular localization with such information, supplemented with additional information (e.g., k-mer frequency). This field has a long history and many prediction tools have been released. Even in this era of proteomic atlas at the single-cell level, researchers continue to develop new algorithms, aiming at accessing the impact of disease-causing mutations/cell type-specific alternative splicing, for example. In this article, we overview the entire field and discuss its future direction.

PSORTm: a bacterial and archaeal protein subcellular localization prediction tool for metagenomics data

Motivation

Many methods for microbial protein subcellular localization (SCL) prediction exist; however, none is readily available for analysis of metagenomic sequence data, despite growing interest from researchers studying microbial communities in humans, agri-food relevant organisms and in other environments (e.g. for identification of cell-surface biomarkers for rapid protein-based diagnostic tests). We wished to also identify new markers of water quality from freshwater samples collected from pristine versus pollution-impacted watersheds.

Results

We report PSORTm, the first bioinformatics tool designed for prediction of diverse bacterial and archaeal protein SCL from metagenomics data. PSORTm incorporates components of PSORTb, one of the most precise and widely used protein SCL predictors, with an automated classification by cell envelope. An evaluation using 5-fold cross-validation with in silico-fragmented sequences with known localization showed that PSORTm maintains PSORTb’s high precision, while sensitivity increases proportionately with metagenomic sequence fragment length. PSORTm’s read-based analysis was similar to PSORTb-based analysis of metagenome-assembled genomes (MAGs); however, the latter requires non-trivial manual classification of each MAG by cell envelope, and cannot make use of unassembled sequences. Analysis of the watershed samples revealed the importance of normalization and identified potential biomarkers of water quality. This method should be useful for examining a wide range of microbial communities, including human microbiomes, and other microbiomes of medical, environmental or industrial importance.

Availability and implementation

Documentation, source code and docker containers are available for running PSORTm locally at https://www.psort.org/psortm/ (freely available, open-source software under GNU General Public License Version 3).

Supplementary information

Supplementary data are available at Bioinformatics online.

Use of Chou's 5-steps rule to predict the subcellular localization of gram-negative and gram-positive bacterial proteins by multi-label learning based on gene ontology annotation and profile alignment

To date, many proteins generated by large-scale genome sequencing projects are still uncharacterized and subject to intensive investigations by both experimental and computational means. Knowledge of protein subcellular localization (SCL) is of key importance for protein function elucidation. However, it remains a challenging task, especially for multiple sites proteins known to shuttle between cell compartments to perform their proper biological functions and proteins which do not have significant homology to proteins of known subcellular locations. Due to their low-cost and reasonable accuracy, machine learning-based methods have gained much attention in this context with the availability of a plethora of biological databases and annotated proteins for analysis and benchmarking. Various predictive models have been proposed to tackle the SCL problem, using different protein sequence features pertaining to the subcellular localization, however, the overwhelming majority of them focuses on single localization and cover very limited cellular locations. The prediction was basically established on sorting signals, amino acids compositions, and homology. To improve the prediction quality, focus is actually on knowledge information extracted from annotation databases, such as protein-protein interactions and Gene Ontology (GO) functional domains annotation which has been recently a widely adopted and essential information for learning systems. To deal with such problem, in the present study, we considered SCL prediction task as a multi-label learning problem and tried to label both single site and multiple sites unannotated bacterial protein sequences by mining proteins homology relationships using both GO terms of protein homologs and PSI-BLAST profiles. The experiments using 5-fold cross-validation tests on the benchmark datasets showed a significant improvement on the results obtained by the proposed consensus multi-label prediction model which discriminates six compartments for Gram-negative and five compartments for Gram-positive bacterial proteins.

Subcellular location prediction of apoptosis proteins using two novel feature extraction methods based on evolutionary information and LDA

Background

Apoptosis, also called programmed cell death, refers to the spontaneous and orderly death of cells controlled by genes in order to maintain a stable internal environment. Identifying the subcellular location of apoptosis proteins is very helpful in understanding the mechanism of apoptosis and designing drugs. Therefore, the subcellular localization of apoptosis proteins has attracted increased attention in computational biology. Effective feature extraction methods play a critical role in predicting the subcellular location of proteins.

Results

In this paper, we proposed two novel feature extraction methods based on evolutionary information. One of the features obtained the evolutionary information via the transition matrix of the consensus sequence (CTM). And the other utilized the evolutionary information from PSSM based on absolute entropy correlation analysis (AECA-PSSM). After fusing the two kinds of features, linear discriminant analysis (LDA) was used to reduce the dimension of the proposed features. Finally, the support vector machine (SVM) was adopted to predict the protein subcellular locations. The proposed CTM-AECA-PSSM-LDA subcellular location prediction method was evaluated using the CL317 dataset and ZW225 dataset. By jackknife test, the overall accuracy was 99.7% (CL317) and 95.6% (ZW225) respectively.

Conclusions

The experimental results show that the proposed method which is hopefully to be a complementary tool for the existing methods of subcellular localization, can effectively extract more abundant features of protein sequence and is feasible in predicting the subcellular location of apoptosis proteins.

An Organelle Correlation-Guided Feature Selection Approach for Classifying Multi-Label Subcellular Bio-Images

Nowadays, with the advances in microscopic imaging, accurate classification of bioimage-based protein subcellular location pattern has attracted as much attention as ever. One of the basic challenging problems is how to select the useful feature components among thousands of potential features to describe the images. This is not an easy task especially considering there is a high ratio of multi-location proteins. Existing feature selection methods seldom take the correlation among different cellular compartments into consideration, and thus may miss some features that will be co-important for several subcellular locations. To deal with this problem, we make use of the important structural correlation among different cellular compartments and propose an organelle structural correlation regularized feature selection method CSF (Common-Sets of Features) in this paper. We formulate the multi-label classification problem by adopting a group-sparsity regularizer to select common subsets of relevant features from different cellular compartments. In addition, we also add a cell structural correlation regularized Laplacian term, which utilizes the prior biological structural information to capture the intrinsic dependency among different cellular compartments. The CSF provides a new feature selection strategy for multi-label bio-image subcellular pattern classifications, and the experimental results also show its superiority when comparing with several existing algorithms.

Protein subcellular localization prediction using multiple kernel learning based support vector machine

An efficient multi-label protein subcellular localization prediction system was developed by introducing multiple kernel learning (MKL) based support vector machine (SVM).