Julian Grandvallet-Contreras

I am a bioinformatics analyst at the University of Colorado Anschutz Medical Campus, embedded in a multi-PI shared resource in the Department of Pediatrics. My work runs across germline and somatic variant analysis, bulk and single-cell RNA-seq, ATAC-seq and CUT&RUN, applied mostly to childhood leukemia, neuro-oncology and cancer evolution.

Most of what I do is building analysis that other people can rely on and reproduce. That means containerized Nextflow pipelines on HPC for six laboratories, written records of the decisions behind duplicate handling and normalization and filtering, and where it matters, a second independent implementation to check the first one. A recent benchmark found that the standard normalization method reports a real fourfold loss of signal as a significant gain, which is the kind of result you only find if you go looking.

I work in R and Python, on nf-core and Nextflow, in the Bioconductor and tidyverse ecosystems. I also teach: an asynchronous R course for researchers, four mentored trainees, and an invited faculty workshop at AACR on using large language models for scientific programming.

Co-author on work in the New England Journal of Medicine, Leukemia, EBioMedicine, Brain Pathology and Blood. Originally from Mexico City, currently in Colorado, and happy to hear from people working on genomics anywhere.