Narendra Solanki

Developed a k-mer-based machine-learning approach for reference-free diagnosis of ADTKD-MUC1, targeting a GC-rich VNTR region. Generated simulated exonic datasets from whole-genome sequencing data to address limited training data. Investigated domain adaptation to improve model generalisation and evaluated diagnostic performance on unseen real short-read sequencing data, achieving comparable performance to reference-based methods, with improvements in sensitivity &/or specificity under selected conditions.

Analysed behavioural and phenotypic datasets from mouse and rat neuroscience studies using R, including data preprocessing, statistical analysis and visualisation. Developed automated, reusable workflows and tools to standardise data processing and improve reproducibility. Managed heterogeneous datasets, addressed data-quality issues and collaborated with researchers to interpret quantitative results and support scientific reporting.

Developed computational workflows for high-throughput time-lapse microscopy of E. coli growth and ageing, including image preprocessing, segmentation, cell tracking and quantitative phenotyping. Adapted and evaluated the DeLTA deep-learning framework for automated cell analysis. Performed quality control and correction of image-analysis artefacts, generating analysis-ready single-cell trajectories and preparing microscopy datasets for downstream analysis.

Developed computational test cases and workflows for bulk and single-cell RNA-seq analysis, covering transcriptomic data processing, statistical analysis and visualisation. Contributed to the design of test cases and task-specific prompts for evaluating LLM-generated R code for transcriptomic data analysis workflows.

Supported the network manager with organisational and administrative tasks. Assisted the scientific coordinator with the preparation and organisation of workshops, events and interviews. Prepared and reviewed reports and supporting documentation.

Investigated visual information processing and neural circuits underlying visually guided behaviour in Drosophila. Redesigned and optimised a flight-simulator system, establishing reproducible behavioural paradigms. Mentored PhD students in thesis writing and research ethics, trained researchers in flight-behaviour techniques, and contributed to researcher development, scientific outreach and institutional events.

Investigated visual learning, novelty recognition and memory in Drosophila using flight-simulator-based behavioural paradigms. Characterised neural circuits involved in novelty choice, focusing on the roles of the central complex and mushroom bodies.

Screened over 1,500 mutant lines to identify genetic candidates affecting locomotor behaviour. Applied behavioural, genetic and imaging techniques to investigate neural circuits underlying locomotion.