TransformerGO: predicting protein–protein interactions by modelling the attention between sets of gene ontology terms

Abstract

Motivation Protein-protein interactions (PPIs) play a key role in diverse biological processes but only a small subset of the interactions have been experimentally identified. Additionally, high-throughput experimental techniques that detect PPIs are known to suffer various limitations such as exaggerated false positives and negatives rates. The semantic similarity derived from the Gene Ontology (GO) annotation is regarded as one of the most powerful indicators for protein interactions. However, while computational approaches for prediction of PPIs have gained popularity in recent years, most methods fail to capture the specificity of GO terms. Results We propose TransformerGO, a model that is capable of capturing the semantic similarity between gene ontology sets dynamically using an attention mechanism. We generate dense graph embeddings for GO terms using an algorithmic framework for learning continuous representations of nodes in networks called node2vec. TransformerGO learns deep semantic relations between annotated terms and can distinguish between negative and positive interactions with high accuracy. TransformerGO outperforms classic semantic similarity measures on gold standard PPI datasets and state-of-the-art machine learning-based approaches on large datasets from S. cerevisiae and H. sapiens. We show how the neural attention mechanism embedded in the transformer architecture detects relevant functional terms when predicting interactions. Availability and implementation https://github.com/Ieremie/TransformerGO Supplementary information Supplementary data are available at Bioinformatics online.

Cracking the black box of deep sequence-based protein-protein interaction prediction

“…we did not need a validation set and added the validation to training for these two datasets. We chose the random 80/20 split since most reviewed methods report the mean accuracy of five-fold cross-validation 4–8,11,18,33,36,37 or a random hold-out test set 2,3,12,13,32,38–41 .…”

Genome-wide characterization leading to simple sequence repeat (SSR) markers development in Shorea robusta

“…The biological process (382), with the highest level of involvement in the functional enrichment analysis results is highlighted here by GO ID and p-value. Lately, this kind of characterization and annotation of genes were used to predict common functions of 12,886 whole-genome duplication (WGD) in S. leprosula (Ng et al 2021 ), examination of differentially expressed genes (Yamasaki et al 2017 ), validation of immune genes (Karthikeyan et al 2021 ), identification of novel prognostic biomarker (Xu et al 2020 ), analyses of Integrated Gene Expression Profiling Data (IGEPA) (You et al 2020 ), identification of the blood-based signatures molecules and drug targets of patients with COVID-19 (Hasan et al 2022 ), and annotation of protein–protein interactions (Ieremie et al 2022 ).…”

TUnA: An uncertainty aware transformer model for sequence-based protein-protein interaction prediction

“…Transformers, widely used in natural language processing for their ability to capture rich long-range dependencies through the multi-headed self-attention mechanism, have shown to be impactful even in drug-protein interaction prediction and protein-protein interaction prediction, as evidenced by TransformerCPI and TransformerGO, respectively. 17–19…”