Predicting response in cancer patients using machine learning models

Current clinical decision making for colorectal cancer (CRC) patients is based on a relatively small number of clinical and pathological hallmarks. Developments in the cost effectiveness and robustness of genome-wide molecular phenotyping greatly expand the number of features that can feed into decision making. The Computational Biology and Integrative Genomics Lab and associated researchers in the Bioinformatics Hub, use machine learning to develop and validate clinical classifiers, based on multi-omic data, which can predict treatment response for cancer patients. If successful, this would allow treatment to be individually tailored, meaning that patients receive the most effective treatment in the first instance.

Recently, Dr. Sanjay Rathee, postdoctoral working under the direction of Prof. Francesca Buffa, developed a model to predict treatment response for CRC patients undergoing chemotherapy, radiotherapy, and oxaliplatin treatment using data from the S:CORT consortium led by Prof. Tim Maughan. Looking at concordant predictions between different methods, this approach addresses the challenge of discovering decision-making genes with minimum noise, ensuring models are general and reliable.

In the first analysis of S:CORT samples (131 rectal cancer samples) from patients treated with radiotherapy and capecitabine, the model gives an accuracy of 90%. Whilst initial accuracy is quite high on retrospective cohorts, the results needs to be validated in external, independent cohorts. If validated, the model could predict response for a new patient based on a pre-treatment biopsy. Finding out whether a treatment is likely to be effective could greatly minimise the number of unnecessary interventions and associated side effects (and cost) whilst also increasing the chance of the chosen intervention being an effective treatment strategy.

Find out more about our research below

The relationship between unexpected weight loss & cancer

New research will help GPs to identify the signs they should look for to swiftly diagnose cancer in people with unexpected weight loss

TP53 gene determines severe blood cancer development

A new study has found that the allelic appearance of the TP53 gene has great impact on development of severe types of blood cancer

Using DNA & RNA to treat cancers

Cancer research UK Oxford Centre Development Fund Awardees Ysobel Baker and Tom Brown investigate the potential of DNA and RNA molecules as precision cancer treatments

Researchers discover cell communication mechanism that drives cancer adaptation

DPAG and Oncology researchers have uncovered a new mechanism by which cancer cells adapt to the stresses they encounter as they grow and respond to therapies

Prof Anna Schuh wins Vice-Chancellor Innovation Award

Anna and her team wins the Teamwork award for their work on improving the outcome of children with blood diseases in sub-Saharan Africa

Using machine-learning approaches to identify blood cancer types

Oxford researchers have outlined the applications of AI in the classification of myeloproliferative neoplasm cancers in a new study

Researchers discover mutation that determines treatment efficiency

Weatherall Institute of Molecular Medicine researchers have recently discovered why a class of cancer drugs is beneficial only in a subset of patients

New digital classification method using AI developed for colorectal cancer

A new study from S:CORT demonstrates an easy, cheap way to determine colorectal cancer molecular subtype using AI deep-learning digital pathology technology

Mapping the T-cell landscape of pancreatic cancer

Through analysis of T-cell populations, researchers Drs Enas Abu-Shah & Shivan Sivakumar identify novel therapeutic opportunities in pancreatic cancer patients