Bioinformatics and Imaging

Professor Peter Hamilton, Lead Investigator

The Cancer Bioinformatics group consists of scientists with expertise across a broad spectrum of subjects including Computational Biology, Computer Vision, Machine Learning, Data Integromics, and Systems and Network Biology. In the era of high-throughput big data generation in genomics and molecular oncology, quantitative computational methods are key for understanding disease processes, identifying druggable targets and developing targeted therapies in cancer.

In complex diseases like cancer, deciphering the vast genomic landscape that underpins the disease presents one of the major challenges in translational and experimental cancer research. Understanding and developing novel computational approaches to interpreting genomic, transcriptomic and proteomic data forms the cornerstone of modern cancer research. Strongly allied to this is the concurrent analysis of tissue phenotypic data. Digital pathology, image analysis and informatics provide important technologies to support high throughput computerised analysis of tissue samples and to measure and validate candidate diagnostic, prognostic and predictive biomarkers in cancer.

 The aim of the group is to lead on the development of novel computational and statistical methods in the analysis of genomic and image data, and to support interdisciplinary collaborative research by working closely together with biologists, pathologists and oncologists within the Centre for Cancer Research and Cell Biology (CCRCB), providing the interface between genomic data, tissue pathology and translational medicine.

 Key research areas include:

  • Computational biology and biostatistics;
  • Pathway analysis, causal inference of regulatory networks, and integration of genetics and genomics data;
  • Tissue imaging, image analysis and tissue biomarker discovery;
  • High-throughput analysis of genomic and image data;
  • Quantitative methods in disease-genes-drugs connection discovery;
  • Data “Integromics”.

Dynamic research within the group spans from basic exploration and method development to applications in specific diseases. Using the high-throughput platforms within the Northern Ireland Molecular Pathology Laboratory (NI-MPL), we generate a wealth of data across a range of cancer specific projects. Using the latest analytics software and in-house algorithms, key hypotheses are generated and tested with the aim of generating novel biomarkers and multivariate signatures for cancer diagnostics and prognostics. Through collaboration with the Northern Ireland Biobank (NIB), NI-MPL and Digital Pathology, the Molecular Pathology Informatics team have also developed a novel “integromics” platform called PICAN (Pathology Integromics in Cancer) for the management of clinical, phenotypic and genotypic data from cancer tissues.

The group has also built up strong expertise and capacity in gene expression connectivity mapping, pushing forward on three fronts in this research field: algorithmic design, novel applications in cancer therapeutics, and high performance software development. The research work of the group has generated a great deal of interest and interactions with QUB colleagues and international collaborators. The recent awards of grants from CR-UK and Leukaemia & Lymphoma NI, both of which use connectivity mapping approach as their essential components, are testimonies of this group’s research impact on academic beneficiaries.

The group has a strong interest in Tissue Imaging and Pathology Informatics and works closely with the NI-MPL and the NIB to achieve these goals. It has one of the most extensive digital pathology laboratories in the UK, with scanning technologies, image and tissue microarray (TMA) management software, along with image analysis capabilities for quantitative biomarker discovery, validation and translation. PathXL Ltd was spun out from these activities and is now a leading digital pathology software company with customers worldwide. It has been working closely with the group on automated tumour identification and has established TissueMark as a leading software platform with dedicated algorithms for tumour markup and analysis.

Finally, the group takes a leading role in the education and mentoring of students and scientists to provide them with a deeper knowledge and understanding of modern methods in bioinformatics, computational biology and tissue imaging, as needed to cope with the data revolution in biology and medicine. This includes a comprehensive MSc course in Bioinformatics and Computational Genomics.