2025
Neuroinformatics
Peer-Reviewed
Functional Connectivity Signatures in fMRI-Derived Connectome for the Diagnosis of Autism Spectrum Disorder
Karim SMS, Rathore RS.
Brain Organoid and Systems Neuroscience Journal, 2025; 3: 170–179. doi:10.1016/j.bosn.2025.06.004 · Cited 3 times
Compared functional connectivity patterns from healthy infants to adults and to ASD patients (age 9.3–35.2 yrs) using graph-based network parameters in fMRI-derived connectomes. Connectome-to-connectome comparison across brain lobes indicates synaptic rewiring concentrated in intra-thalamic and inter-lobe connections with the thalamus, distinguishing developmental from ASD-specific pathological patterns.
2024
Neuroinformatics
Biomedical AI
Peer-Reviewed
Identifying Discriminative Features of Brain Network for Prediction of Alzheimer's Disease Using Graph Theory and Machine Learning
Karim SMS, Fahad MS, Rathore RS.
Frontiers in Neuroinformatics, 2024; 18: 1384720. doi:10.3389/fninf.2024.1384720 · Cited 32 times
Using rs-fMRI data from SALD and OASIS datasets (n=112 participants, age 21–95 yrs), five ML models incorporating graph-theory metrics were developed to predict Alzheimer's disease. 5-fold cross-validation achieved 82–92% accuracy; SVM reached 92%. Analysis identified 13 key discriminating brain regions with significantly reduced connectivity to the thalamus in AD subjects, including substantia nigra and nucleus accumbens.
2023
Biomedical AI
Peer-Reviewed
Prediction of Transient and Permanent Protein Interactions Using AI Methods
Kumar K, Karim SMS, Kumar M, Singh RR.
Bioinformation, 2023; 19(6): 749 · Cited 5 times
43 physicochemical, geometrical and structural features were computed for 402 curated protein-protein complexes (permanent and transient categories). Five supervised ML models (Scikit-learn) plus an ANN (TensorFlow/Keras) were trained to classify interaction type. Models achieved 76.54–82.71% accuracy; k-NN achieved the highest. Provides a computational framework for PPI classification with applications in drug target discovery.
2019
Bioinformatics
Preprint
Understanding the Trends of Antibiotic Resistance Research Through Text Mining and Analysis of PubMed Abstracts
Karim SMS.
2019. Preprint / working paper.
NLP-based retrospective analysis of PubMed abstracts on antimicrobial resistance (AMR) quantified research output by pathogen, author affiliation and clinical trial geography. Found a persistent focus on long-standing pathogens (E. coli, P. aeruginosa) despite international policy frameworks, with AMR research activity concentrated in developed nations and a gap in translating policy into diversified research priorities.
This list is updated periodically. For the most current publications list, including collaborator-authored works that cite 7DNA Analytica, see our LinkedIn page or contact us. Preprint citations are subject to change upon peer-reviewed publication.