Research
Research
A single long read can carry sequence variation, haplotype information, and DNA methylation. I develop computational methods to connect these signals and ask more precise questions about human genomes.
Somatic variation
Cell-type-resolved variation
Bulk sequencing combines DNA from many cell populations. This makes it difficult to determine which cells carry a genetic change, especially for complex structural variants and tandem repeats.
With SniffCell, I use the DNA methylation carried on native long reads to assign informative molecules to methylation-resolvable cell types. This connects variant sequence with cellular context and supports the study of cell-type-associated tandem-repeat variation in human tissues, including the brain.
Genetics & epigenetics
DNA methylation & haplotype phasing
DNA methylation adds a complementary layer of information to the genetic sequence. Long reads let us examine these signals together on individual molecules, rather than treating sequence and epigenetic measurements as separate observations.
I developed MethPhaser to extend haplotype phase blocks using allele-specific methylation, including across regions with limited sequence variation. My work also includes computational approaches for methylation analysis and the integration of genetic and epigenetic information across human populations.
Population epigenomics
All of Us methylation
I study population-scale DNA methylation using long-read sequencing data from the All of Us Research Program. Our study of 1,490 self-identified Hispanic/Latino participants integrates genetic variation, ancestry, and DNA methylation.
The methylation analyses examine allele-specific patterns and associations between tandem-repeat variation and DNA methylation. By measuring genetic and epigenetic information together, this work helps characterize regulatory variation across diverse populations.
Rare disease genomics
Rare-disease genomics
Rare-disease analysis benefits from studying an affected individual alongside both parents. Long reads can resolve small variants, structural changes, repeat expansions, methylation, and inherited haplotypes, but sequencing a family at sufficient depth can be costly.
Our Trio-barcoded Adaptive Sequencing (TBAS) workflow combines barcoding with phenotype-driven target selection to sequence a trio on one nanopore flow cell. This work connects method development with the practical challenges of investigating unresolved rare diseases.
Algorithms & benchmarking
Algorithms & benchmarking
My background is in computer science, with an emphasis on alignment, genome assembly, and structural-variant detection. Vulcan combines complementary alignment strategies to improve long-read mapping and variant calling.
Through collaborative work on somatic-mutation benchmarking, cancer genomics, transcriptomics, and metagenomics, I help evaluate how computational methods perform across different samples, technologies, and biological questions.