From biology to public health bioinformatics with StaPH-B
Lee Katz and Andrew Page talk with Curtis Kapsak and Kevin Libuit about moving from biology degrees at James Madison University into public health bioinformatics and contracting. They discuss research mentorship, APHL/CDC fellowships, quality control and the StaPH-B community. Their experiences show how collaborative networks and shared computing resources can help laboratories develop analytical capacity, while Kevin also reflects on making room for life outside work.
Listen to Episode 58: StaPH-B careers with Curtis Kapsak and Kevin Libuit on SoundCloud
In this episode
Finding research beyond medicine
Both guests entered biology with medicine in mind. Kevin joined James Herrick’s research laboratory at James Madison University to strengthen his medical school application, but discovered that he enjoyed contributing his own questions and results to science. Curtis likewise moved away from pre-med, having initially been inspired by Hugh Laurie’s character in House MD. They valued research’s collaborative spirit, symbolised by the purple-and-gold tie-dyed laboratory coats they wore and later handed down to the university.
Their work on environmental antimicrobial resistance and plasmids led from observable resistance phenotypes to questions about genotype. An Ion Torrent PGM and participation in Oxford Nanopore’s early access programme gave them sequencing data to analyse. Stephen Turner mentored them by sharing analyses and scripts that they tried to reproduce. Kevin particularly liked computational work’s quicker feedback loops and the ability to scale what they built.
Fellowships and a community of mentors
Kevin followed an internship at Virginia’s public health laboratory with an APHL/CDC bioinformatics fellowship. Curtis took the same fellowship route at Colorado’s state laboratory. They describe a one- or two-year placement at a state laboratory or CDC laboratory, with projects through which fellows develop practical bioinformatics skills. For Curtis, this marked the change from mixed wet-laboratory and computational research to full-time bioinformatics.
Curtis spent eight months in Colorado, supported by a mentor and the previous bioinformatics fellow, before joining Lee’s CDC team as a contractor. Kevin contrasts the limited established mentorship available to his earlier cohort with the support network that subsequently developed. StaPH-B’s Slack channel connected scientists across state laboratories and CDC, providing a community beyond any individual host laboratory.
Quality before interpretation
Kevin identifies three major challenges when he entered public health bioinformatics in 2016: computing infrastructure, workforce development and getting analytical tools running. Colleagues also needed to understand what sequencing could contribute to public health decisions. At Virginia’s DCLS, his initial priority was quality control: confidence in reported data had to come before more elaborate biological interpretation because those results could inform government action and intervention.
Curtis describes starting with enteric foodborne pathogens in Colorado, particularly Salmonella enterica and E. coli. Other PulseNet organisms included Listeria, Campylobacter and Shigella. He also worked on healthcare-associated pathogens, including Klebsiella and Citrobacter. His subsequent CDC work focused on enteric pathogens; by this recording, his new contracting role was focused on SARS-CoV-2.
Remote work and a changed sabbatical
Kevin’s first public health opportunity came through his supervisor’s professional network rather than a deliberate search for that sector. Because the work was remote and used AWS cloud computing, he could begin supporting Virginia’s laboratory while travelling in Bogotá, Colombia. He connects that early experience with his later ability to support international partners through cloud resources.
In 2019, Kevin and his wife planned a self-funded international sabbatical for 2020, with activities in Spain, Vietnam and Argentina. The pandemic changed those plans. He stayed at DCLS until around July, developing SARS-CoV-2 resources and hiring and training his replacement. They then took three months travelling domestically, backpacking and camping. Although leaving a lead scientist position felt risky, he says the break clarified his priorities and did not harm his career.
Contracting and shared cloud workflows
Moving into contracting broadened the range of laboratories Kevin could support beyond Virginia, including work with APHL and CDC Thailand. He argues that the influx of SARS-CoV-2 funding should also strengthen capacity for other infectious diseases rather than produce resources useful only during the pandemic.
He describes three enabling developments: Docker containers, workflow languages and web applications connecting those workflows to cloud computing. Terra, from the Broad Institute, provides the interface between workflows and Google Cloud resources. A workflow built once can then serve multiple laboratories, shifting much of the support effort towards configuring inputs, understanding outputs and training users. Shared videos and openly available code help extend that support beyond one-to-one assistance, while keeping technical dependency management away from users who need to concentrate on public health results.
From sequencing reads to submitted genomes
Curtis describes inputs consisting of Illumina or Nanopore FASTQ files and sample metadata. In Terra, laboratories can import a CSV table pointing to data in Google storage buckets and launch an analysis. Only five days into his new role, he had already seen laboratories move from available sequencing data to submission-ready SARS-CoV-2 genomes within an hour; this was an observed example, not a promise for every laboratory.
Kevin discusses ARTIC V3 tiled-amplicon reads, genome assembly, lineage and clade typing, and preparing metadata for GISAID or GenBank. Submission itself remains a substantial barrier. Andrew notes that even COG-UK’s centralised arrangements require ongoing contact with archives to resolve problems. The wider goal is to make these capabilities available to laboratories without requiring each to recruit a highly experienced bioinformatician.
Highlights
- 00:02:34 — Kevin explains how pre-med ambitions turned into a commitment to biological research.
- 00:05:24 — Research collaboration leads into sequencing, shared scripts and computational work.
- 00:11:28 — APHL/CDC fellowships provide routes into full-time public health bioinformatics.
- 00:14:07 — Kevin contrasts early fellowship mentorship with the growing StaPH-B support network.
- 00:15:04 — Infrastructure, workforce development and analytical work depend on reliable quality control.
- 00:17:08 — Curtis traces his work from foodborne and healthcare-associated pathogens to SARS-CoV-2.
- 00:18:58 — Professional connections and remote cloud computing bring Kevin into public health.
- 00:22:14 — A planned international sabbatical becomes a three-month domestic break during the pandemic.
- 00:28:22 — Contracting expands partnerships and creates opportunities beyond SARS-CoV-2.
- 00:29:54 — Docker, workflow languages and Terra make support for many laboratories more practical.
- 00:32:08 — Curtis describes FASTQ and metadata inputs and rapid laboratory onboarding.
- 00:35:23 — International capacity building and open resources create a need for more bioinformatics talent.
In their own words
So I just prefer the quicker feedback loops and being able to also scale what we were building.
— Kevin Libuit, 00:05:24
And not to say, obviously, quality is of utmost importance in every field, but in public health, it informs action, government action and intervention.
— Kevin Libuit, 00:15:04
Who is talking
- Lee Katz (host)
- Andrew Page (host)
- Curtis Kapsak (guest, Theiagen Genomics)
- Kevin Libuit (guest, Theiagen Genomics)
Also mentioned: James Herrick, Stephen Turner.
Tools and resources mentioned
Ion Torrent PGM, Docker, Terra, AWS, Google Cloud, Slack, GitHub, BaseSpace, GISAID, GenBank, ARTIC V3 tiled-amplicon sequencing.
Questions this episode answers
How did Curtis Kapsak and Kevin Libuit enter public health bioinformatics?
Both moved from pre-med interests into biological research at James Madison University, then developed sequencing and computational skills during their studies. APHL/CDC bioinformatics fellowships placed Kevin at Virginia’s public health laboratory and Curtis at Colorado’s state laboratory.
Why did Kevin prioritise quality control at DCLS?
Public health results can inform government action and intervention, so the laboratory needed confidence in the data it reported. Kevin describes quality control as an early priority before expanding into genetic relatedness, genomic epidemiology and antimicrobial resistance detection.
How did Terra help them support laboratories analysing SARS-CoV-2?
Terra connected shared workflows to Google Cloud resources through a web interface. Laboratories could supply sequencing reads and metadata, while training focused on configuring inputs and interpreting outputs rather than installing and maintaining every analytical dependency.
Did Kevin’s sabbatical damage his bioinformatics career?
Kevin says it did not negatively affect his career opportunities. After helping prepare DCLS for his departure, he took a three-month domestic break with his wife and returned with clearer personal and professional priorities.
This page was written from a machine transcript of the episode and checked against it. Listen to the episode for the whole conversation.
Show notes
We dive deeper into Curtis’s and Kevin’s career! Relevant links: https://www.skypeascientist.com/ https://www.aphl.org/fellowships/Pages/Bioinformatics.aspx https://www.jmu.edu/genomics/index.shtml jmu.edu/biology