Cambridge Cancer Genomics

Overview

Cambridge Cancer Genomics, or CCG.ai, built software to help oncologists choose and monitor cancer treatment from genomic and clinical data. Its premise was that a tumour changes under treatment, so one genomic snapshot is not enough. Repeated analysis could reveal response or resistance and support a change in therapy.

The company joined Y Combinator's Summer 2017 batch, partnered with Genomics England, and developed the OncOS machine-learning platform. Dante Labs acquired CCG.ai in June 2021 and integrated OncOS with its genomics-interpretation software. Dante Labs

Founding Story

John Cassidy, Nirmesh Patel, and Evaline Tsai founded CCG.ai in Cambridge in 2016. The team combined cancer biology, genomics, and machine learning around a clinical question: which drug should a patient receive now, and when has the tumour changed enough to require another strategy?

The company built software rather than a single laboratory assay. That choice let it combine tumour sequencing, inherited genomic data, liquid biopsies, and medical context. The ambition was to develop biomarkers of treatment response from longitudinal evidence rather than rely only on broad population averages.

Timeline

• 2016: CCG.ai is founded and participates in Accelerate Cambridge.
• 2017: The company joins YC's Summer batch.
• 2019: CCG.ai and Genomics England announce an immunotherapy partnership using 100,000 Genomes Project data.
• 2021: Dante Labs acquires CCG.ai in June and announces integration of OncOS into Immensa.

What They Built

OncOS analyzed medical, germline, somatic, and liquid-biopsy data for precision oncology. Liquid biopsies can sample tumour DNA from blood, offering a less invasive way to observe change across treatment. Machine-learning models and biomarkers aimed to identify whether therapy was working or resistance was emerging.

The Genomics England project focused on immunotherapy in lung and colorectal cancer. It developed a sequencing panel for tumour mutational burden plus genes associated with response and resistance. The plan was to pair the panel with liquid-biopsy analysis so clinicians could monitor a tumour over time. Genomics England partnership

Market Position

Target Customers

CCG.ai targeted oncology clinics, research groups, drug developers, and genomic-testing organizations. The clinical user was an oncologist, but adoption depended on laboratories, validated workflows, regulators, health systems, and reimbursement.

Market Size

Precision oncology is large because ineffective cancer treatment is costly and harmful. Yet a headline market estimate does not translate directly into software revenue. Each use requires evidence for a particular cancer, therapy, specimen, and decision point.

Competition

CCG.ai competed with genomic interpretation vendors, liquid-biopsy companies, hospital molecular tumour boards, and internal pharmaceutical biomarker teams. Its distinction was the longitudinal treatment-response layer. Its constraint was the need for clinical validation and integration into regulated testing.

Business Model

Possible buyers could pay for interpretation software, testing workflows, research collaborations, or biomarker-development work. Public sources do not provide audited revenue, pricing, margins, or patient volumes. Acquisition terms were also undisclosed.

Traction

The Genomics England partnership gave CCG.ai access to a serious national genomics program and a defined immunotherapy research question. Dante Labs' decision to integrate OncOS into its CE-IVD, ISO 13485 Immensa software provides evidence that the buyer valued the technology. It does not by itself establish clinical utility across cancers.

Post-Mortem

Dante Labs acquired CCG.ai in June 2021. The buyer wanted one platform connecting genomic testing with oncology interpretation. CCG.ai supplied machine-learning software and treatment-response work; Dante Labs supplied sequencing, diagnostics operations, and a commercial genomics platform. Accelerate Cambridge independently records the buyer and date. Accelerate Cambridge

The combination addressed a common precision-medicine bottleneck. An algorithm can be scientifically interesting but hard to deploy without sample logistics, quality systems, validated interpretation, and clinical distribution. Joining a diagnostics company gave CCG.ai a path from research software toward a delivered test.