Jagath Chandana Rajapakse

This paper introduces BRAINMAP, a model for ADHD diagnosis using dynamic functional connectivity, achieving improved detection with three fMRI datasets.

This paper introduces MHASNet, a deep learning model for taste assessment using EEG signals, achieving high accuracy across five taste categories.

This paper introduces PharmacoBridge, a diffusion bridge-based method for de novo drug design guided by pharmacophore arrangements, generating molecules with high binding affinity to target proteins.

This paper proposes a sequential modelling approach for medical image segmentation using a modified MegaByte model, achieving comparable performance on intracranial haemorrhage segmentation and outperforming existing methods on tiny structures.

This paper introduces BrainGAT, a graph neural network for gender classification using multimodal brain connectomes, achieving 83.13% accuracy on the Human Connectome Project dataset.

DrDimont predicts differential drug response by integrating multi-omics networks and provides explainable results, outperforming methods based on differential protein expression or PageRank in breast cancer data.

This paper proposes a method to simplify deep neural networks for decoding brain functional connectomes, achieving similar performance with 2% of the original parameters and identifying relevant biomarkers.

The paper introduces a modular refinement algorithm to detect smaller, functionally significant modules in protein-protein interaction networks that classical methods miss.

This paper develops a deep neural network approach for predicting neuroblastoma patient outcomes using graph-based feature extraction, achieving higher accuracy than traditional models.

The paper uses deep neural networks to classify brain states in AD, MCI, and CN subjects using fMRI data and identifies salient brain regions associated with these conditions.

MultiDCoX identifies differentially co-expressed gene sets influenced by multiple factors in transcriptomic data, demonstrating effectiveness through simulated data and breast cancer analysis.

The paper introduces a method that improves functional similarity computation by incorporating GO term enrichment, enhancing performance in predicting sequence homologies, gene expression correlations, and protein-protein interactions.