Research and Projects

Measurable residual disease (MRD) monitoring is critical for assessing treatment response and predicting relapse in pediatric acute lymphoblastic leukemia (ALL). Current MRD protocols rely on patient-specific immunoglobulin (Ig) and T-cell receptor (TCR) gene rearrangements as molecular markers. However, approximately 5-10% of ALL patients lack informative Ig/TCR markers, leaving a gap in personalized monitoring strategies.

KMT2A gene fusions, offer an alternative marker for these patients. These chromosomal rearrangements create unique fusion junctions that can serve as patient-specific molecular signatures for detecting minimal residual leukaemic cells.

In this ongoing project, I'm developing a bioinformatics pipeline to identify KMT2A gene fusions from RNA-seq data and integrate them into the targeted next-generation sequencing (NGS) workflow for MRD monitoring.

The advent of protein structure prediction software like AlphaFold has revolutionised structural biology, generating an unprecedented volume of protein structures now accessible in public databases. This wealth of data opens new avenues for discovery across protein science, particularly in remote homology detection, where structural information can reveal distant evolutionary relationships that elude traditional sequence-based methods.

When sequence similarity drops into the "midnight zone" (typically below 20-25% identity), conventional sequence-based approaches struggle to identify homologous proteins. However, protein structure is far more conserved than sequence throughout evolution, making structure-based comparisons invaluable for detecting these relationships.

In this project, I developed a computational workflow using Python and structural bioinformatics tools to detect, cluster, and visualise distantly related viral annealase protein families by structural similarity.