Projects available

E1: Combining CRISPR-screens and AI to Identify Novel Therapeutic Strategies to Target Cancer-Stroma Interactions 

​Applications will be reviewed on a rolling-basis up until the deadline of July 31st 2024.  The studentship made be awarded prior to the July deadline and those interested in applying are encouraged to submit their application as soon as it's ready.

As tumours progress, the relationship between cancer cells and the tumour microenvironment (TME) evolves to promote growth and evade anti-tumour immunity. This multidisciplinary project will combine the fields of cancer biology, bioinformatics and artificial intelligence to identify tractable targets for manipulating the cancer-TME interaction for clinical translation. This requires the development of new methodologies capable of modelling fundamental biological mechanisms." 

During tumorigenesis, malignant phenotypes evolve along with molecular susceptibilities that can be targeted by therapies. Initiatives like the cancer dependency map (DepMap) aim to systematically unravel these vulnerabilities through extensive loss-of-function (CRISPR) screens across cancer cell lines. However, these studies are not performed in a microenvironment that is representative of a patient cancer and focus solely on cancer cells themselves. This project will address this limitation by exploring strategies to target the interactions between cancer, stromal and immune cells.  

Working with colleagues from the BCI and Exscientia, a leading AI drug discovery company, you will integrate in-house and publicly available genome-wide CRISPR screen, proteomic and transcriptomic datasets. To delineate novel targetable pro-tumourigenic pathways, you will apply cutting-edge machine learning methods to construct interpretable models of cancer-stroma-immune interactions. Your computational predictions will be tested by colleagues in the Cameron laboratory in state-of-the-art 3D heterotypic co-cultures that accurately mimic TMEs. Results from lab experiments will feed a prediction-validation learning loop, ensuring the continual improvement of computational models and resulting target predictions. Promising new targets with the potential to impede tumour progression and manipulate the immune suppressive microenvironment will be exploited in concert with Exscientia’s AI-driven target identification, prioritization and drug design capabilities.  

We are seeking a highly motivated candidate eager to apply computational methods to complex biological questions. The ideal candidate will bridge cancer biology and data science, with a solid understanding of molecular biology and some experience performing data wrangling, statistics and/or machine learning in a modern programming language (e.g., R/Python). The successful candidate will benefit from mentorship and training provided by the BCI and Exscientia, and gain exposure to modern AI-enhanced oncology research and target identification efforts. 

Eligibility and Applying 

This project is part of the UKRI/BBSRC AI for Drug Discovery Progamme, and successful candidates will join a cohort of students working on complementary projects in the AI for Drug Discovery space.

We are looking for highly motivated individuals who are passionate about contributing to new discoveries in drug discovery bioscience through the application of the latest techniques in AI and data science. Ideal candidates will have a grounding in both a natural science and data science, e.g. through a Master's degree or work experience in a subject such as bioinformatics or computational chemistry. Alternatively, you may have, for example, a first-class degree in computer science followed by biochemistry experience, or vice versa (qualifications and evidence thereof must be obtained before October 2024). You will be confident in performing data wrangling and analysis in a language such as Python, R or C++. Effective communication skills are essential.

We particularly encourage students from groups that are currently underrepresented in postgraduate science research, including black and minority ethnic students and those from a socio-economically disadvantaged background.

This studentship award can cover UK tuition fees, UKRI stipend (currently £20,622) and a consumable allowance for a period of 4-years (pro-rata for part-time). Candidates must meet the UKRI requirements to be eligible for the award. Typically this means candidates have unrestricted access on how long they can remain in the UK (i.e. are a British National, have settled, or pre-settled status, have indefinite leave to remain etc.) and have been living in the UK for the 3 years immediately prior to studentship starting.

Visit the Apply Pages for details on how to submit an application. 

Supervisors:
Dr Otto Morris - Senior Biological Data Scientist, Exscientia

Dr Angus Cameron - Reader in Cell Signalling and Tumour Cell Biology, arts Cancer Institute, QMUL

Professor Pedro Cutillas - Professor of Cell Signalling and Proteomics, QMUL, Barts Cancer Institute, QMUL

Project Partner:
Exscientia 

F1: AI-based Identification of New Drug Targets for Personalised Oncology 

​Applications will be reviewed on a rolling-basis up until the deadline of July 31st 2024.  The studentship made be awarded prior to the July deadline and those interested in applying are encouraged to submit their application as soon as it's ready.

The genes of cancer cells acquire somatic alterations at a rate higher than non-cancer cells. Some of these altered genes are essential to drive the different phases of cancer evolution and are therefore called cancer drivers. Identifying cancer drivers is one of the major goals of cancer biology because their knowledge highly improves our understanding of cancer mechanisms and expands the repertoire of possible therapeutic interventions. Cancer drivers are however challenging to find because they are only a small minority of all cancer altered genes. Recently, our group has developed a machine learning method for cancer driver identification in individual patients. Our approach is first trained to learn the features of a well characterised set of cancer drivers. It then predicts as cancer drivers the altered genes in individual patients that best resemble these features. We have applied our approach to almost 10,000 cancer samples from 34 cancer types, building a vast repertoire of cancer drivers. In this project we aim to refine this knowledge towards the prediction of new drug targets for personalised oncology. The student will first extract useful information on druggability as well as known drug targets from curated resources including DrugBank, GDSC, Meta-Knowledgebase, DGIdb, STICH, CTD. They will then build a new classifier aimed to predict new potential targets among all available cancer drivers. Predictions will include either the repurposing of already available drugs as anticancer agents in specific contexts, or the identification of new targets for the development of new drugs. 

The ideal candidate will have a background in computer science or statistics. Some knowledge in artificial intelligence classification and/or cancer biology will be favourable.

Eligibility and Applying 

This project is part of the UKRI/BBSRC AI for Drug Discovery Progamme, and successful candidates will join a cohort of students working on complementary projects in the AI for Drug Discovery space.

We are looking for highly motivated individuals who are passionate about contributing to new discoveries in drug discovery bioscience through the application of the latest techniques in AI and data science. Ideal candidates will have a grounding in both a natural science and data science, e.g. through a Master's degree or work experience in a subject such as bioinformatics or computational chemistry. Alternatively, you may have, for example, a first-class degree in computer science followed by biochemistry experience, or vice versa (qualifications and evidence thereof must be obtained before October 2024). You will be confident in performing data wrangling and analysis in a language such as Python, R or C++. Effective communication skills are essential.

We particularly encourage students from groups that are currently underrepresented in postgraduate science research, including black and minority ethnic students and those from a socio-economically disadvantaged background.

This studentship award will cover a UKRI Stipend (currently £20,622), a consumable allowance, and UK tuition fees only, for a period of 4 years (pro-rate for part-time study). To be classed as home for fee purposes, candidates typically need to have unrestricted access on how long they can remain in the UK (i.e. are a British National, have settled, or pre-settled status, have indefinite leave to remain etc.) and have been living in the UK for the 3 years immediately prior to studentship starting. Fee status is determined by the Admissions department at the point of application. Candidates who would be classed as overseas are welcome to apply for this opportunity, however must state in their application other funding they have secured to cover the difference in tuition fees.

Visit the Apply Pages for further details on how to submit an application. 

Supervisors:
Francesca Ciccarelli - Lead of the Centre for Cancer Genomics and Computational Biology, Bart's Cancer Institute

M2: Developing a High-Quality PPI Interactome to Model Cell States and Disease Biology

Application Deadline: July 31st 2024

Modelling biological function is a difficult challenge. There are about 20,000 protein coding elements in the human genome which translate to over 220,000 protein coding isoforms. There are also ~45,000 non coding regulatory elements in the human genome. These elements interact inside cells to produce various outcomes. In disease, some of these interactions fail or go outside the normal range, thus leading to disruption of normal cell function. This can lead to activation of various compensatory mechanisms which may alleviate the problem, provide partial resolution, or make the problem worse. The ability to model such activity could go a long way towards providing novel avenues to better understand cell function and finding treatments for disease.

This challenge is confounded by the high levels of noise in biological data. Some of the data we have lacks true connections and has false connections that are due to experiential noise. In addition, we often know that two elements interact, but not the direction or sign of the interaction. This noise and lack of detail limits the types of analysis that can be performed. There is also a lack of context as many interactions are derived from experimental models that do not represent specific cells. With good quality data, it would be possible to model outcomes of activation or inhibition of a gene on the other interactors and build more reliable link prediction tools.

There are several methods in the field that attempt to do this work already, so the first step will be to review and collate a database of gold standard positive and negative interactions and create a framework for benchmarking PPI quality. The second step is to review and create a method for inferring directionality and sign of PPI interactions. This data can then be used to train various models to infer the likelihood of a given edge in the graph. Finally the directed and signed high confidence interactome can then be used to predict biological assay readouts, perturb-seq gene signatures and generally attempt to model cell biology and infer key regulators of disease states.

High level objectives:

• Create a framework and database to benchmark edge quality in knowledge graphs
• Use additional omics & regulatory information to infer sign and direction of PPI and other interactions
• Then use this data to improve edge prediction and removal models to assign confidence to known edges and suggest missing ones
• Use directed and signed interactome to model pathway activity in cells and predict modulators for drug discovery and causal reasoning

This project is in conjunction with MSD.

Eligibility and Applying 

This project is part of the UKRI/BBSRC AI for Drug Discovery Progamme, and successful candidates will join a cohort of students working on complementary projects in the AI for Drug Discovery space.

We are looking for highly motivated individuals who are passionate about contributing to new discoveries in drug discovery bioscience through the application of the latest techniques in AI and data science. Ideal candidates will have a grounding in both a natural science and data science, e.g. through a Master's degree or work experience in a subject such as bioinformatics or computational chemistry. Alternatively, you may have, for example, a first-class degree in computer science followed by biochemistry experience, or vice versa (qualifications and evidence thereof must be obtained before October 2024). You will be confident in performing data wrangling and analysis in a language such as Python, R or C++. Effective communication skills are essential.

We particularly encourage students from groups that are currently underrepresented in postgraduate science research, including black and minority ethnic students and those from a socio-economically disadvantaged background.

Candidates must meet the UKRI requirements to be eligible for the award. Typically this means candidates have unrestricted access on how long they can remain in the UK (i.e. are a British National, have settled, or pre-settled status, have indefinite leave to remain etc.) and have been living in the UK for the 3 years immediately prior to studentship starting.

Visit the Apply Pages for further details on how to submit an application. 

Supervisors:  
Professor Pedro Cutillas, Professor of Cell Signalling and Proteomics, Barts Cancer Institute at Queen Mary University London 

Dr Kirill Shkura, Systems Biologist and Data Scientist, MSD

Project Partner:  
MSD