Biology subjects

Rajapakse, J. C.

Publications and source records attributed to Rajapakse, J. C..

12 recordsLinked to original sources

bioRxiv2025

Encoding and Decoding of Brain Dynamic Functional Connectivity for ADHD Diagnosis

Recent studies have demonstrated strong associations between the changes in dynamic functional connectivity (FC) and both behavioral and cognitive functions. The sliding window technique is the most widely used method for evaluating dynamic FC; however, it faces two key challenges: distributional shifts across windows and high dimensionality, as FC is computed across windows of the entire time series. To address these issues, we propose BRAINMAP (Bi-level Representation using Attention for INterpretability with Mamba-Aided Prediction) to model the dynamic FC of the brain. BRAINMAP employs the Optimal Transport technique to correct distributional shifts across sliding windows and leverages Graph Neural Networks (GNNs) in conjunction with a hybrid approach that integrates an attention mechanism and the Mamba block to effectively capture spatiotemporal features for functional MR images. Finally, we introduce a novel Top-K sliding window feature selection algorithm to induce sparsity in dynamic FC. We conducted an extensive evaluation of our model for diagnosing Attention Deficit Hyperactivity Disorder (ADHD) using three resting-state fMRI datasets: ADHD-200, UCLA, and CNI-TLC, which comprise a total of 447 subjects with ADHD and 845 typically developing controls. Our architecture outperformed existing state-of-the-art dynamic FC models in ADHD detection, achieving improvements ranging from 3% to 12% across the three datasets. We demonstrate that our proposed model produces robust biomarkers, most notably the connection between the dorsal attention network and the visual network. Using an association study, we further establish the clinical relevance of the identified biomarkers.

neuroscience↗

bioRxiv2025

An EEG-based Multi-Scale Hybrid Attention and Squeeze Network for objective taste assessment

Taste perception is central to flavor experiences. Electroencephalography (EEG) signals carry rich information about taste perception. Because these neural data are independent of verbal reports and bypassing conscious filtering, EEG holds promise as a neural-sensing platform for objective taste representation. However, taste-evoked EEG signals exhibit complex spatiotemporal and spectral dynamics that require advanced computational approaches to decode effectively. This study developed an EEG-based Multi-Scale Hybrid Attention and Squeeze Network (MHASNet) for taste evaluation. EEG signals evoked by distinct taste stimuli were recorded, a dedicated taste-EEG dataset was compiled, and a novel deep learning architecture, MHASNet, was designed to classify these signals. MHASNet synergistically integrates multi-scale convolutions to capture temporal dynamics across different time scales, dual-attention mechanisms to localize discriminative brain regions and electrode positions, and a squeeze-and-excitation module to optimize frequency-band contributions-collectively enabling precise extraction of taste-specific neural signatures. Results showed that the proposed model achieved superior performance across five taste categories (sweet, sour, salty, bitter, and tasteless), with 94.33% accuracy, 91.37% F1-score, 91.89% precision, and 92.07% recall. These results surpass benchmark models while maintaining millisecond-level inference latency suitable for real-time applications. By complementing subjective evaluations and instrumental analyses, the model offers an objective, neurophysiology-based solution for taste evaluation.

neuroscience↗

bioRxiv2024

Pharmacophore-constrained de novo drug design with diffusion bridge

De novo design of bioactive drug molecules with potential to treat desired biological targets is a profound task in the drug discovery process. Existing approaches tend to leverage the pocket structure of the target protein to condition the molecule generation. However, even the pocket area of the target protein may contain redundant information since not all atoms in the pocket is responsible for the interaction with the ligand. In this work, we propose PharmacoBridge, a phamacophore-guided de novo design approach to generate drug candidates inducing desired bioactivity via diffusion bridge. Our method adapts the diffusion bridge to effectively convert pharmacophore arrangements in the spatial space into molecular structures under the manner of SE(3) -equivariant transformation, providing sophisticated control over optimal biochemical feature arrangements on the generated molecules. PharmacoBridge is demonstrated to generate hit candidates that exhibit high binding affinity with potential protein targets.

bioinformatics↗

bioRxiv2023

Predicting gender from structural and functional connectomes via brain and population graph neural networks

Gender differences in terms of structural and functional organization of the human brain have been extensively studied, but existing works have mostly been limited to single modalities. In this paper, we propose a graph attention network architecture (BrainGAT) that uses informative subjectlevel features extracted from multimodal brain graphs to construct a population graph for gender classification. We show that while the extracted subject-level features can be directly used for classification, using these graph embeddings to construct a population graph further improves model performance. On the gender classification task, BrainGAT outperforms baseline models and existing multimodal modeling approaches, achieving an accuracy of 83.13% on the Human Connectome Project dataset. Salient connections highlighted by BrainGAT include connections between the inferior parietal and dorsolateral prefrontal areas of the cortex for females, while connections within the posterior cingulate cortex are highly salient for males. In sum, BrainGAT enables multimodal data to be modelled via population graphs in a parameter-efficient way. Clinical RelevanceSeveral neurological conditions exhibit significant differences across genders and multimodal studies on these diseases are increasingly prevalent. This work highlights gender differences of multimodal connectomes in neurotypical settings. These insights could help to separate multimodal disease biomarkers from fundamental gender differences.

neuroscience↗

bioRxiv2023

Sequential Modelling for Medical Image Segmentation

An image can be seen as a long sequence of tokens with spatial structure hence an image segmentation task can be treated as a sequence-to-sequence task. Existing attention-based segmentation models have incorporated an up-sampling module at the pixel decoder, such design assumed the backbone to have multiple feature scales. Furthermore, the existing models did not emphasize the ability of sequential modelling for multi-tasking. In this paper, we propose to model image segmentation task as a long sequence prediction task with an improvisation on MegaByte. As MegaByte efficiently reduced the self-attention cost using multi-scale structure known as global and local models, we train two MegaByte models as an encoder and decoder for multi-tasking. Our multi-task model allows different segmentation tasks to be trained simultaneously and inferenced with a task prompt. The encoder takes in the image sequence to encode them as a key while the decoder takes in the conditional input as a query. As decoder conditional input is usually sparser than the image, we use a relatively shallow decoder. We demonstrate our method on segmentation of intracranial haemorrhages from computed tomography (CT) head scans and show comparable performance to existing models. On a brain tissue segmentation task on CT images, our method outperformed the existing methods especially on tiny structure. Clinical relevanceCT images are the first line of head scans when stroke or haemorrhage is suspected. Segmentation of intracranial haemorrhages (ICH) can provide clinicians with important measures of brain lesions to decide on treatment procedure or surgical decision. Brain and tissue volumes are indicative of neurodegenerative diseases. The proposed techniques can be used to segment ICH and brain tissues from CT head scans.

neuroscience↗

bioRxiv2022

DrDimont: Explainable drug response predictionfrom differential analysis of multi-omics networks

MotivationWhile it has been well established that drugs affect and help patients differently, personalized drug response predictions remain challenging. Solutions based on single omics measurements have been proposed, and networks provide means to incorporate molecular interactions into reasoning. However, how to integrate the wealth of information contained in multiple omics layers still poses a complex problem. ResultsWe present DrDimont, Drug response prediction from Differential analysis of multi-omics networks. It allows for comparative conclusions between two conditions and translates them into differential drug response predictions. DrDimont focuses on molecular interactions. It establishes condition-specific networks from correlation within an omics layer that are then reduced and combined into heterogeneous, multi-omics molecular networks. A novel semi-local, path-based integration step ensures integrative conclusions. Differential predictions are derived from comparing the condition-specific integrated networks. DrDimonts predictions are explainable, i.e., molecular differences that are the source of high differential drug scores can be retrieved. We predict differential drug response in breast cancer using transcriptomics, proteomics, phosphosite, and metabolomics measurements and contrast estrogen receptor positive and receptor negative patients. DrDimont performs better than drug prediction based on differential protein expression or PageRank when evaluating it on ground truth data from cancer cell lines. We find proteomic and phosphosite layers to carry most information for distinguishing drug response. AvailabilityDrDimont is available on CRAN: https://cran.r-project.org/package=DrDimont. Contactkatharina.baum@hpi.de

bioinformatics↗

bioRxiv2020

Obtaining leaner deep neural networks for decoding brain functional connectome in a single shot

Neuroscientific knowledge points to the presence of redundancy in the correlations of brains functional activity. These redundancies can be removed to mitigate the problem of overfitting when deep neural network (DNN) models are used to classify neuroimaging datasets. We propose an algorithm that removes insignificant nodes of DNNs in a layerwise manner and then adds a subset of correlated features in a single shot. When performing experiments with functional MRI datasets for classifying patients from healthy controls, we were able to obtain simpler and more generalizable DNNs. The obtained DNNs maintained a similar performance as the full network with only around 2% of the initial trainable parameters. Further, we used the trained network to identify salient brain regions and connections from functional connectome for multiple brain disorders. The identified biomarkers were found to closely correspond to previously known disease biomarkers. The proposed methods have cross-modal applications in obtaining leaner DNNs that seem to fit the data better. The corresponding code is available at https://github.com/SCSE-Biomedical-Computing-Group/LEAN_CLIP.

bioinformatics↗

bioRxiv2019

Refining modules to determine functionally significant clusters in molecular networks

Module detection algorithms relying on modularity maximization suffer from an inherent resolution limit that hinders detection of small topological modules, especially in molecular networks where most biological processes are believed to form small and compact communities. We propose a novel modular refinement approach that helps finding functionally significant modules of molecular networks. The module refinement algorithm improves the quality of topological modules in protein-protein interaction networks by finding biologically functionally significant modules. The algorithm is based on the fact that functional modules in biology do not necessarily represent those corresponding to maximum modularity. Larger modules corresponding to maximal modularity are incrementally re-modularized again under specific constraints so that smaller yet topologically and biologically valid modules are recovered. We show improvement in quality and functional coverage of modules using experiments on synthetic and real protein-protein interaction networks. Results were also compared with six existing methods available for clustering biological networks. In conclusion, the proposed algorithm finds smaller but functionally relevant modules that are undetected by classical quality maximization approaches for modular detection. The refinement procedure helps to detect more functionally enriched modules in protein-protein interaction networks, which are also more coherent with functionally characterised gene sets.

bioinformatics↗

bioRxiv2019

A deep neural network approach to predicting clinical outcomes of neuroblastoma patients

The availability of high-throughput omics datasets from large patient cohorts has allowed the development of methods that aim at predicting patient clinical outcomes, such as survival and disease recurrence. Such methods are also important to better understand the biological mechanisms underlying disease etiology and development, as well as treatment responses. Recently, different predictive models, relying on distinct algorithms (including Support Vector Machines and Random Forests) have been investigated. In this context, deep learning strategies are of special interest due to their demonstrated superior performance over a wide range of problems and datasets. One of the main challenges of such strategies is the "small n large p" problem. Indeed, omics datasets typically consist of small numbers of samples and large numbers of features relative to typical deep learning datasets. Neural networks usually tackle this problem through feature selection or by including additional constraints during the learning process. We propose to tackle this problem with a novel strategy that relies on a graph-based method for feature extraction, coupled with a deep neural network for clinical outcome prediction. The omics data are first represented as graphs whose nodes represent patients, and edges represent correlations between the patients omics profiles. Topological features, such as centralities, are then extracted from these graphs for every node. Lastly, these features are used as input to train and test various classifiers. We apply this strategy to four neuroblastoma datasets and observe that models based on neural networks are more accurate than state of the art models (DNN: 85%-87%, SVM/RF: 75%-82%). We explore how different parameters and configurations are selected in order to overcome the effects of the small data problem as well as the curse of dimensionality. Our results indicate that the deep neural networks capture complex features in the data that help predicting patient clinical outcomes.

bioinformatics↗

bioRxiv2019

Decoding brain functional connectivity implicated in AD and MCI

Deep neural networks have been demonstrated to extract high level features from neuroimaging data when classifying brain states. Identifying salient features characterizing brain states further refines the focus of clinicians and allows design of better diagnostic systems. We demonstrate this while performing classification of resting-state functional magnetic resonance imaging (fMRI) scans of patients suffering from Alzheimers Disease (AD) and Mild Cognitive Impairment (MCI), and Cognitively Normal (CN) subjects from the Alzheimers Disease Neuroimaging Initiative (ADNI). We use a 5-layer feed-forward deep neural network (DNN) to derive relevance scores of input features and show that an empirically selected subset of features improves accuracy scores for patient classification. The common distinctive salient brain regions were in the uncus and medial temporal lobe which closely correspond with previous studies. The proposed methods have cross-modal applications with several neuropsychiatric disorders.

neuroscience↗

bioRxiv2017

Combining Semantic Similarity and GO Enrichment for Computation of Functional Similarity

Functional similarity between genes is widely used in many bioinformatics applications including detecting molecular pathways, finding co-expressed genes, predicting protein-protein interactions, and prioritization of candidate genes. Methods evaluating functional similarity of genes are mostly based on semantic similarity of gene ontology (GO) terms. Though there are hundreds of functional similarity measures available in the literature, none of them considers the enrichment of the GO terms by the querying gene pair. We propose a novel method to incorporate GO enrichment into the existing functional similarity measures. Our experiments show that the inclusion of gene enrichment significantly improves the performance of 44 widely used functional similarity measures, especially in the prediction of sequence homologies, gene expression correlations, and protein-protein interactions.\n\nSoftware availabilityThe software (python code) and all the benchmark datasets evaluation (R script) are available at https://gitlab.com/liuwt/EnrichFunSim.

bioinformatics↗

bioRxiv2017

MultiDCoX: Multi-factor Analysis of Differential Co-expression

BackgroundDifferential co-expression signifies change in degree of co-expression of a set of genes among different biological conditions. It has been used to identify differential co-expression networks or interactomes. Many algorithms have been developed for single-factor differential co-expression analysis and applied in a variety of studies. However, in many studies, the samples are characterized by multiple factors such as genetic markers, clinical variables and treatments. No algorithm or methodology is available for multi-factor analysis of differential co-expression.\n\nResultsWe developed a novel formulation and a computationally efficient greedy search algorithm called MultiDCoX to perform multi-factor differential co-expression analysis of transcriptomic data. Simulated data analysis demonstrates that the algorithm can effectively elicit differentially co-expressed (DCX) gene sets and quantify the influence of each factor on co-expression. MultiDCoX analysis of a breast cancer dataset identified interesting biologically meaningful differentially coexpressed (DCX) gene sets along with genetic and clinical factors that influenced the respective differential co-expression.\n\nConclusionsMultiDCoX is a space and time efficient procedure to identify differentially co-expressed gene sets and successfully identify influence of individual factors on differential co-expression.\n\nSoftwareR function will be available upon request.

bioinformatics↗