Senior Bioinformatics Scientist

Cubiq Recruitment; Greater London; Hybrid; GBP 117,000 - 143,000; Full time

Job summary

Cubiq Recruitment is assisting a London-based BioAI company in recruiting a Senior Bioinformatics Scientist to shape data strategy for genomic and protein foundation models. You will curate and evaluate training data, and align model objectives with biological questions.

The role involves collaborating with ML engineers and biologists, designing novel architectures for genomics, and guiding data-centric model development. Hybrid London setup with substantial impact.

Qualifications

• Built novel deep learning architectures or frameworks applied to genomics or protein modelling.
• Hands-on experience with Transformer-based or foundation model architectures in a biological context (e.g. Enformer, Nucleotide Transformer, Evo, HyenaDNA, ESM, or comparable).
• Understanding of regulatory genomics, non-coding variant prediction, or gene regulation for modelling decisions informed by biology.
• Experience curating or evaluating training data for large-scale biological models.
• Deep learning must be a core competency; biology alone is not sufficient, and general ML experience without biological applications won’t translate.

Responsibilities

• Shape the data strategy behind genomic and protein models; decide data needs and curate training datasets.
• Fine-tune models for real biological applications and collaborate with ML engineers and biologists to translate questions into model objectives.
• Translate complex biological questions into concrete model targets and ensure alignment with research aims.

Skills

Deep learning

Genomics

Transformer models

Data curation

Model fine-tuning

Job description

Senior Bioinformatics Scientist (Genomic Foundation Models)

London (Hybrid) | Up to £130,000 + Equity | Some flexibility for exceptional candidates

We're partnering with a London-based BioAI company that is doing something genuinely rare.

Backed by one of the world's leading AI infrastructure companies and working alongside frontier AI partners, they've built the largest proprietary biological dataset on the planet, spanning billions of genes across millions of species. Following a recent Series B, they're using that data to train genomic and protein foundation models capable of designing entirely novel biological systems.

Their ambition is to compress decades of biological discovery into years. The dataset they're training on doesn't exist anywhere else, which means the modelling opportunities here are unlike anything available in academia or at other companies working off public data.

The Role

You’ll sit at the centre of their foundation model efforts, shaping the data strategy behind their genomic and protein models. You’ll be deciding what data the models need, curating and evaluating training datasets, fine-tuning models for real biological applications, and working directly with ML engineers and biologists to translate biological questions into model objectives.

Requirements

• You have built novel deep learning architectures or frameworks applied to genomics or protein modelling. This is the defining requirement for this role. We’re looking for people who design and implement models, not solely use or fine-tune existing ones.
• Hands‑on experience with Transformer‑based or foundation model architectures in a biological context (e.g. Enformer, Nucleotide Transformer, Evo, HyenaDNA, ESM, or comparable)
• Understanding of regulatory genomics, non-coding variant prediction, or gene regulation at a level where you can make modelling decisions informed by the biology
• Experience curating or evaluating training data for large‑scale biological models
• Deep learning must be a core competency, not a secondary skill. Strong biology alone is not sufficient, and general ML experience without biological applications won’t translate.

Strong Preferences

• Published work applying deep learning to functional genomics or variant effect prediction
• Experience with protein language models (ProGen, ESMFold, or similar)
• Familiarity with models like DeepSEA, Sei, Basenji, BPNet, ChromBPNet, SpliceAI, or AlphaGenome