Computational Scientist / Senior Computational Scientist

Switchpoint Bio is decoding the molecular switchpoints that drive human health and disease. The choices our cells make — which genes to activate, how to respond to environment and exposure, when to mount immune defenses — are governed by epigenetic mechanisms. We bring together genomics, epigenomics, and AI to turn epigenetic biology into decision-grade evidence for drug discovery and clinical decision-making.

Epigenetics is how cells record genetics, aging, environment, and exposure across a lifetime — and it shapes nearly every major area of human disease. Our flagship programs focus on autoimmune conditions and cancer today, but the same mechanisms underlie cardiovascular, metabolic, and neurodegenerative disease. Together, these touch billions of lives — yet most treatments are still chosen by trial and error. We're building the platform that makes these mechanisms visible — and the evidence base that turns it into better drugs, earlier diagnoses, and treatments matched to the person, not the population.

Founded by pioneers in single-cell biology, epigenetics, and genomics, we operate across Helsinki and Boston. We're a small founding team in the earliest days — building not just a platform, but a company and a culture from scratch. We hire builders: people who take ownership, move fast, and thrive when the problem is bigger than the playbook.

About the role

We have an established team building machine learning models and infrastructure. For this role, we're looking for the scientist who brings the deep domain knowledge to point these tools towards new biological opportunities.

As a Computational Scientist you'll be the biological and analytical lead on our multiomic data: designing how we integrate genomics, epigenomics, transcriptomics, and proteomics with biobank cohorts, health registries, and exposure data, and interpreting what it means for disease mechanism and drug discovery. You'll work closely with our ML scientists — you shape the scientific questions, the study design, the ground truth, and the biological validation; together you turn high-dimensional molecular data into evidence the company can act on.

Responsibilities

• Lead the design and analysis of multimodal data integration — combining molecular data (genomics, epigenomics, transcriptomics, proteomics), single-cell and biobank-scale cohorts, health registries, and lifestyle/exposure data into coherent, interpretable analyses.
• Bring domain judgment to modeling: define the biological questions, choose the right cohorts and comparisons, engineer biologically meaningful features, and decide what "good" looks like for immune cell-type identification, disease classification, and treatment-response prediction.
• Own biological interpretation and validation — connect molecular signals to mechanisms (immune regulation, epigenetic control, disease pathways), and separate real biology from batch effects, technical artifacts, and confounding.
• Read the frontier of the genomics, epigenomics and immunology literature independently, judge what's credible, and recommend what we should build or adopt next.
• Partner with the ML scientist to turn scientific questions into models and evals, and translate model output back into decision-grade biological evidence.
• Work with wet-lab and clinical collaborators to shape assay and study design so the resulting data is analyzable, well-controlled, and answers the question we actually care about.

What you'll help build in your first 12 months

• Multiomic integration framework — establish how we bring molecular, cohort, registry, and exposure data together, with the QC, harmonization, and confounder handling that make cross-cohort analysis trustworthy.
• Immune cell-type and disease signatures — define and validate the biological ground truth behind our immune cell-type identification and disease classifiers, so the models the team ships reflect real biology.
• Mechanistic evidence for programs — deliver analyses that link epigenetic and molecular signals to disease mechanisms in autoimmune conditions and cancer, feeding target and biomarker decisions.
• Clinical and translational read-outs — help build predictors of health outcomes and treatment response from high-dimensional molecular profiles, and own the question of whether they generalize across populations.
• A shared scientific workflow — with the ML scientist and agentic tooling, shape how analyses, hypotheses, and literature synthesis flow through the team so science moves faster without losing rigor.

Requirements

Must-have

• Multiomics domain expertise — you've worked hands-on with omics data (e.g. genomics, epigenomics, transcriptomics, methylation, proteomics) and understand the biology behind it, not just the matrices. Single-cell and/or biobank-scale experience is strongly preferred.
• Multimodal / health data integration — you've integrated and analyzed heterogeneous biological and health data, and you know how to make signals from different modalities and cohorts comparable.
• Independent scientific judgment — you read cutting-edge papers in genomics/genetics/epigenomics, assess what holds up, and recommend concrete solutions.
• Quantitative and computational rigor — solid statistics and hands-on Python (or R) for real analysis; you don't need to be an ML infrastructure engineer, but you collaborate fluently with one.
• Fluency with how naive analysis fails on molecular data — batch effects, technical artifacts, ascertainment, population structure, and confounding; you anticipate them and design around them.
• Clear cross-disciplinary communication — you translate between biology and ML/engineering, explain your reasoning, and write and speak fluent English.
• High-impact published science — a track record of strong, high-impact publications is especially favored.
• PhD in relevant field.

Nice to have

• Domain depth in immunology, or autoimmune disease biology.
• Experience with epigenomic assays and data (ATAC-seq, DNA methylation, single-cell multiomics).
• Familiarity with genomic/epigenomic foundation models (e.g., AlphaGenome, Evo, Nucleotide Transformer, scGPT, Geneformer) and how to use them for real biological questions.
• Experience with causal inference, Mendelian randomization, or QTL integration for target and biomarker validation.
• Prior work in startups, frontier labs, or research environments where you owned problems end-to-end.
• Experience working in distributed, cross-timezone teams.

Benefits

• Generous paid time off.
• Health and Dental insurance.
• 401k with matching.
• Life and short/long-term disability insurance.
• MBTA commuter benefits.

Pay

$140,000–200,000/year depending on the qualifications + meaningful option package. As a first-wave hire, your option package reflects the impact you'll have on what we build.

Schedule

Hybrid 2-3 days/week at a company office in Cambridge, MA, USA.

Why join us

• Be the scientific voice that decides what we model and whether it's real — at the frontier of genomics, epigenetics, and human health.
• Own the biology end-to-end, from study design and integration to interpretation and translational impact.
• Work in a true complementary pair with a strong ML scientist, supported by experienced computational scientists, rather than doing everything alone.
• Help build a company, platform, and culture from the ground up, with high ownership and real influence from day one.
• Collaborate with globally recognized scientists across Helsinki and Boston.
• Solve meaningful problems with real-world clinical impact, on a lean team that moves fast and builds ambitiously.

Relocation support is available for strong candidates moving to Boston. Occasional travel is required, primarily to our Helsinki hub, plus the occasional conference (a handful of trips per year). Reports to the Chief Scientific Officer.