Biology subjects

Hoque, M. T.

Publications and source records attributed to Hoque, M. T..

3 recordsLinked to original sources

bioRxiv2026

ASD-FER: An Annotated Multi-Class Benchmark and Comparative Study for Facial Emotion Recognition in Children with Autism Spectrum Disorder

Autism Spectrum Disorder (ASD) affects social interaction and communication, and automatically recognizing emotions from the facial expressions of children with ASD remains challenging. A key barrier to facial emotion recognition (FER) in children with ASD is the absence of ASD-specific datasets with proper emotion-level annotations. To address this, we present ASD-FER, a benchmark of 2,779 facial images of children with ASD annotated across six emotion categories by three independent annotators using majority voting. We evaluate nine models -- CNN, SVM, AlexNet, ResNet34, ResNet50, MobileNetV2, and three backbone configurations of our proposed task-specific architecture, EQFormer: ViT, Swin-T, and a compact Hybrid CNN-Transformer (HybridT) -- under stratified 10-fold cross-validation and a held-out 90/10 evaluation. ResNet50 led conventional models at 84.89% mean cross-validation accuracy; among transformer-based models, ViT led with 88.27%, followed by Swin-T at 88.02% and HybridT at 86.58%. On the held-out set, ViT achieved the highest overall accuracy (91.01%), while Swin-T obtained the best balanced accuracy (74.47%) and macro-F1 (0.7078); HybridT remained competitive at 89.57% accuracy with a fraction of the parameters. To support explainability, Grad-CAM visualizations highlighted attention on emotion-relevant facial regions -- eyes, eyebrows, and mouth -- with Swin-T the most spatially focused and the HybridT the most parameter-efficient (2.66M). Class imbalance and cross-dataset generalization remain open challenges.

neuroscience↗

bioRxiv2026

ESMDisPred: A Structure-Aware CNN-Transformer Architecture for Intrinsically Disordered Protein Prediction

Intrinsically disordered proteins (IDPs) lack stable three-dimensional structures, yet play vital roles in key biological processes, including signaling, transcription regulation, and molecular scaffolding. Their structural flexibility presents significant challenges for experimental characterization and contributes to diseases such as cancer and neurodegenerative disorders. Accurate computational prediction of IDPs is important for advancing research and drug discovery, structural biology, and protein engineering. In this study, we introduce ESMDisPred, a novel structure-aware disorder predictor that builds on the representational power of Evolutionary Scale Modeling-2 (ESM2) protein language models. ESMDisPred integrates sequence embeddings with structural information from the Protein Data Bank (PDB) to deliver state-of-the-art prediction accuracy. Model performance is further enhanced through feature engineering strategies, including terminal residue encoding, statistical summarization, and sliding-window analysis. To capture both local sequence motifs and long-range dependencies, we designed a hybrid CNN-Transformer architecture that balances convolutional efficiency with the representational power of self-attention. On CAID3 benchmarks, our latest model achieves ROC-AUC 0.895, AP 0.778, and a max F1 of 0.759, outperforming recent methods. Our results highlight the importance of integrating protein language model embeddings with explicit structural information for improved disorder prediction.

bioinformatics↗

bioRxiv2020

AIRBP: Accurate identification of RNA-binding proteins using machine learning techniques

MotivationIdentification of RNA-binding proteins (RBPs) that bind to ribonucleic acid molecules, is an important problem in Computational Biology and Bioinformatics. It becomes indispensable to identify RBPs as they play crucial roles in post-transcriptional control of RNAs and RNA metabolism as well as have diverse roles in various biological processes such as splicing, mRNA stabilization, mRNA localization, and translation, RNA synthesis, folding-unfolding, modification, processing, and degradation. The existing experimental techniques for identifying RBPs are time-consuming and expensive. Therefore, identifying RBPs directly from the sequence using computational methods can be useful to efficiently annotate RBPs and assist the experimental design. In this work, we present a method, called AIRBP, which is designed using an advanced machine learning technique, called stacking, to effectively predict RBPs by utilizing features extracted from evolutionary information, physiochemical properties, and disordered properties. Moreover, our method, AIRBP is trained on the useful feature-subset identified by the evolutionary algorithm (EA). ResultsThe results show that AIRBP attains Accuracy (ACC), F1-score, and MCC of 95.38%, 0.917, and 0.885, respectively, based on the benchmark dataset, using 10-fold cross-validation (CV). Further evaluation of AIRBP on independent test set reveals that it achieves ACC, F1-score, and MCC of 93.04%, 0.943, and 0.855, for Human test set; 91.60%, 0.942 and 0.789 for S. cerevisiae test set; and 91.67%, 0.953 and 0.594 for A. thaliana test set, respectively. These results indicate that AIRBP outperforms the current state-of-the-art method. Therefore, the proposed top-performing AIRBP can be useful for accurate identification and annotation of RBPs directly from the sequence and help gain valuable insight to treat critical diseases. AvailabilityCode-data is available here: http://cs.uno.edu/~tamjid/Software/AIRBP/code_data.zip

bioinformatics↗