Colm J. Ryan | UCD Cancer Data Lab

The synthetic lethal interaction between CDS1 and CDS2 is a vulnerability in uveal melanoma and across multiple tumor types

Tue, 01 Jul 2025 00:00:00 +0000

Proteomic compensation by paralogs preserves protein interaction networks after gene loss in cancer

Wed, 28 May 2025 15:19:22 +0000

Hamda's paper evaluating genetic interaction scoring systems for combinatorial CRISPR screens

Tue, 08 Apr 2025 00:00:00 +0000

Congrats to Hamda on her first publication with the lab, Benchmarking Genetic Interaction Scoring Methods for Identifying Synthetic Lethality from Combinatorial CRISPR Screens, up now on biorXiv.

Combinatorial CRISPR screens, where two genes are perturbed simultaneously, are increasingly used to identify new synthetic lethal targets in cancer. They’ve been particularly popular for identifying synthetic lethal relationships between paralog pairs. In order to identify synthetic lethal hits from these screens it is necessary to identify and quantify genetic interactions – instances where the effect observed from perturbing two genes diverges from what is expected based on perturbing each gene individually. Multiple scoring pipelines for quantifying genetic interactions from these screens have been published, but comparison across pipelines has been limited (in part due to the lack of an external benchmark). Indeed most published screens use distinct scoring systems, making their results difficult to compare.

In this work we propose that paralog synthetic lethal interactions identified from DepMap single-gene CRISPR screens can be used as an external benchmark. These pairs have been found to be synthetic lethal across cell lines from diverse cancer types by associating the loss of one gene with increased sensitivity to inhibition of another. They are therefore at least somewhat robust to genetic background and unlikely to be cell line specific. This is not a perfect benchmark, as some pairs that are synthetic lethal in only specific contexts will be missed, but overall gene pairs found to be synthetic in a panel of cell lines should get a higher score than those where there is no evidence of synthetic lethality in the panel.

To compare scoring systems using this approach, Hamda rescored 5 published combinatorial CRISPR datasets (total 21 screens) using 5 distinct scoring systems. Getting the scoring systems to work across multiple screens and rescoring everything was a lot of work but allows us to compare scoring systems in a systematic fashion. We found that no single method performs best across all screens but identify two methods that perform well across most datasets.

You can read the full paper here for all of the comparisons!

Benchmarking Genetic Interaction Scoring Methods for Identifying Synthetic Lethality from Combinatorial CRISPR Screens

Tue, 01 Apr 2025 00:00:00 +0000

Congratulations to Dr. Anjan Venkatesh

Wed, 19 Feb 2025 00:00:00 +0000

Congratulations to Dr. Anjan Venkatesh, who successfully defended his thesis ‘A systematic study of paralog and protein complex response to gene loss in cancer’ on February 14th (💘).

Anjan has been a member of the group since 2020, joining us from the excellent SFI Centre for Research Training in Genomics Data Science. Prior to joining our lab, Anjan worked with Prof. Ken Wolfe (perhaps surprisinglym not on paralogs!). You can read the first part of Anjan’s thesis on biorxiv. The second part is still in the oven.

Many thanks to the external examiner, Prof. Pedro Beltrao, and internal examiner, Prof. Denis Shields. Thanks also to Dr. Dimitris Chatzopoulos for chairing the viva and making sure everything ran smoothly.

Anjan and Colm

Congratulations to Dr. Cosmin Tudose

Wed, 20 Nov 2024 00:00:00 +0000

Congratulations to Dr. Cosmin Tudose, who successfully defended his thesis ‘Defining molecular vulnerabilities in childhood leukaemia through biological network analysis’ on November 19th.

Cosmin has been a member of the Bond and Ryan groups since 2020, joining us from the excellent SFI Centre for Research Training in Genomics Data Science. You can read the first part of Cosmin’s thesis, published in NAR Cancer, here. The second part is still in the oven.

Many thanks to the external examiner, Dr. Lisa Russell , and internal examiner, Prof. David MacHugh. Thanks also to Dr. David Gomez for chairing the viva and making sure everything ran smoothly.

Olivier's paper evaluating alternative protein representations for predicting shared paralog functions

Wed, 16 Oct 2024 00:00:00 +0000

Congrats to Olivier on his first publication with the lab, Evaluating Sequence and Structural Similarity Metrics for Predicting Shared Paralog Functions, up now on biorXiv.

Comparisons between protein sequences are the foundation of much of modern biology, allowing us to infer evolutionary relationships between genes and assign functions to homologous proteins. These sequence comparisons are not just made across species, for inferring functions of orthologs, but also within species, for paralogs (our favourites!). For decades, protein sequence identity has been used as a proxy for the functional similarity of paralogs. However, the advent of alternative AI-derived protein representations (predicted structures from AlphaFold or embeddings from protein language models) raises the possibility that similarity metrics derived from these alternative representations might capture functional relationships that sequence identity misses. Here we set out to test this systematically – Can similarity between AlphaFold predicted structures better identify paralogs with shared functions than simple sequence identity? What about protein language model embeddings?

We evaluated two definitions of shared paralog functions (shared protein-protein interaction partners, synthetic lethality) across two species (budding yeast and human). We found that, at least for some tasks, similarity measures derived from alternative protein representations can outperform sequence identity alone (e.g. similarity of predicted AlphaFold structures better predicts shared protein-protein interactions in yeast). However, more importantly, these measures were not redundant with sequence identity, i.e. combining them with sequence identity led to improved predictions across all tasks. We show that the complementary nature of these alternative metrics applies to other definitions of shared functionality for paralogs (shared gene ontology terms). Overall it is clear that similarity metrics from these alternative protein representations can capture aspects of shared paralog functionality that are not captured by sequence identity alone.

The shared functions of paralogs are relatively easy to assess (e.g. through double perturbation experiments) and make them ideal candidates for the sort of systematic evaluation we have performed here. Looking beyond paralogs, our results suggest that combining these alternative protein representations—sequence, structure, and language model embeddings—could be highly valuable for inferring common functions in a more general sense, including for orthologs.

You can read the full paper here.

We make the full set of similarity metrics available for all yeast and human paralog pairs here. We are happy to generate them for other species too – if you are interested, send us a mail.

Anjan's paper on proteomic compensation between paralogs in cancer

Tue, 15 Oct 2024 00:00:00 +0000

Compensation between duplicate genes (paralogs) allows cancer cells to tolerate genetic mutations that might otherwise be lethal. We wanted to know what this compensation looked at a molecular level – what are the proteins involved doing? How come losing functional genes doesn’t just just break the protein-protein interaction networks operating inside cancer cells?

We used proteomic profiling of isogenic knockouts (in collaboration with Alexander Von Kriegsheim) and large-scale analysis of tumour proteomes to explore this question systematically. We found tons of examples of proteomic compensation, where loss of one gene was associated with increased abundance of its paralog, as well as collateral loss, where loss of one gene was associated with decreased abundance of its paralog. Proteomic compensation pairs were more central in the protein interaction network, more evolutionary conserved, and more likely to be synthetic lethal.

Overall our results support a model where loss of an important gene in cancer (one that has lots of interactions, is involved in an essential function) triggers an increase in the protein abundance of its paralog. Tumour cells then become dependent on the upregulated paralog for survival.

You can read the full paper here.

You can read Anjan’s tweetorial about the paper below:

Very pleased to see my PhD work out as a preprint in bioRxiv! Here, we set out to answer a basic question— how come gene loss in cancer doesn't break protein-protein interaction networks? (1/17) pic.twitter.com/OAqK1b6BMh
— Anjan Venkatesh (@AnjanVenkatesh) October 3, 2024

Evaluating Sequence and Structural Similarity Metrics for Predicting Shared Paralog Functions

Mon, 14 Oct 2024 00:00:00 +0000

Conway Festival 2024!

Thu, 10 Oct 2024 00:00:00 +0000

The lab was well represented at the Conway Festival of Research 2024. Colm gave an invite lectured, Cosmin was a runner up in the Gold Medal competition for his NAR Cancer paper and Narod, Metin, Olivier and Hamda all presented posters. We didn’t take enough photos, but managed to capture Metin and Olivier in action.

Metin presenting his poster at the Conway Festival of Research 2024

Olivier presenting his poster at the Conway Festival of Research 2024.

Congratulations to Dr. Swathi Upadhya

Wed, 13 Dec 2023 00:00:00 +0000

Congratulations to Dr. Swathi Upadhya, who successfully defended her thesis ‘On the predictability and prediction of protein abundances in tumours’ on December 12th.

Swathi has been a member of the lab since 2018! You can read about some of Swathi’s published papers here and here. Swathi also won Best Student Talk at the Irish Mass Spectrometry Society Conference 2022, which you can read about here ` Many thanks to the external examiner, Prof. Oliver Kohlbacher and internal examiner, Assoc. Prof. Dere Greene. Thanks also to Dr. Andrew Hines for chairing the viva and making sure everything ran smoothly.

Swathi with the lab after her viva

Complex synthetic lethality in cancer

Thu, 30 Nov 2023 00:00:00 +0000

Gene essentiality in cancer is better predicted by mRNA abundance than by gene regulatory network-inferred activity

Tue, 28 Nov 2023 00:00:00 +0000

Exploiting Cancer Synthetic Lethality in Cancer---Lessons Learnt from PARP Inhibitors

Sat, 18 Nov 2023 00:00:00 +0000

Barbara's paper on homozygous deletions of paralogs in cancer

Tue, 14 Nov 2023 00:00:00 +0000

Congrats to Dr. Barbara De Kegel on the publication of her work to understand how the dispensability of paralogs shapes homozygous deletion patterns in tumors.

Tumors tolerate astonishing numbers of genetic alterations, including the homozygous deletion of entire protein coding genes. In this work we explore one potential explanation – paralogs! The paper is out in its final form in Molecular Systems Biology here.

You can read the original tweetorial about the preprint below:

the latest work from @dekegelb. Tumors tolerate astonishing numbers of genetic alterations, including homozygous deletion of entire protein coding genes. This suggests significant genetic robustness https://t.co/TQWYcrlp0e 1/11
— Colm Ryan (@colmr) June 27, 2022

Paralog dispensability shapes homozygous deletion patterns in tumor genomes

Tue, 14 Nov 2023 00:00:00 +0000

Swathi's paper on the influence of antibody quality on mRNA-protein correlations

Thu, 18 May 2023 00:00:00 +0000

Congrats to Swathi on the publication of her work to understand the influence of antibody quality on the mRNA-protein correlations observed in tumour cohorts.

A quarter of the antibodies used in The Cancer Genome Atlas (TCGA) RPPA profiles are labelled ‘Use with Caution’, but their protein measurements are still widely used in systematic analyses. In this new paper we show that measurements from these antibodies result in lower mRNA-protein correlations and poorer agreement with mass-spectrometry quantification.

This paper is part of Swathi’s ongoing efforts to understand the how the observed relationships between mRNA and protein abundances are influenced by technical factors.

The paper is out in its final form in Life Science Alliance here. If you read the earlier preprint the primary addition is a new section where we attempt to predict which proteins will be labelled ‘Use with Caution’.

You can read Swathi’s tweetorial about the paper below:

Glad to see that our paper is now out @LSAjournal ! In this work, we analysed how antibody reliability influences the correlation between mRNA and protein abundances for proteins profiled using Reverse Phase Protein Arrays (RPPA)https://t.co/6d0RJ0zbmJ
— Swathi R Upadhya (@swathirupadhya) May 15, 2023

Genetic interactions under the microscope

Wed, 17 May 2023 00:00:00 +0000

Antibody reliability influences observed mRNA–protein correlations in tumour samples

Thu, 11 May 2023 00:00:00 +0000

Congratulations to Dr. Thomas Lefeivre

Thu, 27 Apr 2023 00:00:00 +0000

Congratulations to Dr. Thomas Lefeivre, who successfully defended his thesis ‘Investigation of the functional interplay between signalling pathways and epigenetic factors in acute leukaemia’ on April 18th.

Thomas has been a member of the Bond and Ryan groups since 2018. We’re lucky to have Thomas stay on with us temporarily, experimentally some testing some new synthetic lethals and performing proteomic profiling of cancer cell lines.

Many thanks to the external examiner, Dr. David O’Connor, and internal examiner, Assoc. Prof. Rory Johnson. Thanks also to Dr. Melinda Halasz for chairing the viva and making sure everything ran smoothly.

Thomas after his Viva, dressed as his Irish alter-ego Tomás MacGabhann

Cosmin's paper on using Gene Regulatory Networks to predict gene essentiality in cancer

Wed, 08 Mar 2023 00:00:00 +0000

Gene regulatory networks (GRNs) are widely used to understand how driver mutations in tumours alter transcriptional networks and to identify new therapeutic targets. Here, we use genome-wide CRISPR screen data to systematically evaluate the ability of GRNs to predict sensitivity to genetic perturbation in cancer.

Spoiler alert – we find that, for the purposes of predicting gene inhibition sensitivity, simple mRNA abundance generally outperforms gene regulatory activity inferred from GRNs!

You can read the full paper here.

You can read Cosmin’s tweetorial about the paper below:

Gene regulatory networks (GRNs) are used to identify vulnerabilities in cancer, but it's unclear how good they are at this. Here, we used CRISPR screens to systematically evaluate the ability of GRNs to identify genetic dependencies in cancer. https://t.co/T3DaGm6t6t Thread: 1/16
— Cosmin Tudose (@cosmint_) March 6, 2023

Targeting synthetic lethal paralogs in cancer

Mon, 06 Mar 2023 00:00:00 +0000

We are hiring!

Thu, 17 Nov 2022 00:00:00 +0000

We are currently looking for computational postdoctoral fellows! More details here

Photo above is from a lab hike during COVID times. In Dublin we cannot promise constant sunshine, but we can promise the benefits of living in a capital city coupled with close proximity to the sea and some lovely scenery.

A more recent hike in Howth

The return of in person conferences!

Thu, 03 Nov 2022 00:00:00 +0000

After more than two years with conference travel halted due to COVID-19, lab members have finally had the opportunity to attend in person conferences.

Two years is a long time in the course of a four-year PhD, so it’s been really great to see everyone get the chance to visit new places, present their work, and chat to other scientists.

Although I’m delighted that in person conferences have returned, I’m also really glad that hybrid conferences seem to be sticking. Cosmin and I recently attended #CRISPR22 in person at the Wellcome Genome Campus in Hinxton while other lab members were able to attend online. Aside from the financial and environmental advantages resulting from virtual attendance this also means that visa applications are not required for every conference.

Things kicked off gently in June with a day trip to the University of Limerick for VIBE 2022 (the Virtual Institute of Bioinformatics and Evolution All-Ireland Conference) where Cosmin, Anjan and Swathi all gave talks.

Anjan presented at the British Society for Proteome Research Annual Conference 2022 in Oxford in July.

Anjan presenting at the British Society for Proteome Research Annual Conference 2022

Swathi presented at the 2022 International Mass Spectrometry Conference in Maastricht in August.

Swathi presenting at the 2022 International Mass Spectrometry Conference. Photo courtesy of Pedro Beltrao.

Cosmin presented at CRISPR and Beyond: Perturbations at Scale to Understand Genomes at the Wellcome Genome Campus in Hinxton in September.

Cosmin presenting at CRISPR and Beyond: Perturbations at Scale to Understand Genomes

Thomas presented at the President’s Symposium for the best abstracts at the Haematology Association of Ireland Annual Conference in Cork in October.

Thomas presenting at the President’s Symposium for the best abstracts at the Haematology Association of Ireland Annual Conference. Photo courtesy of Jonathan Bond

Narod is part of the organising committee for the Irish Computational Biology and Genomics Symposium which takes place in person in Galway in December. There’s still a chance to register if you want to attend!

Swathi's paper on the relationship between mRNA and protein abundances

Thu, 27 Oct 2022 00:00:00 +0000

Everyone knows that there is an imperfect correlation between mRNA and protein abundances. In Swathi’s recent paper we analysed how much measurement (ir)reproducibility contributes to this.

We found that some proteins are more reproducibly quantified than others and that in general reproducibly quantified proteins tend to have higher mRNA-protein correlations. Overall variation in measurement reproducibility accounts for a substantial fraction of the unexplained variation between mRNA and protein abundances. You can read the full paper here.

As part of the work we have put together an aggregated protein reproducibility score that we think will be useful for prioritising proteins whose abundance can be reproducibility quantified.

Swathi has also put together a short video presentation of the work for JRNcLub here.

You can read also read the original tweetorial about the paper below:

Great to see this work from @swathirupadhya out in @CellRepMethods! Everyone knows that there is an imperfect correlation between mRNA and protein abundances. In this work we analysed how much measurement (ir)reproducibility contributes to this. 🧵https://t.co/8XBsW61Qd8
— Colm Ryan (@colmr) September 13, 2022

Lab move to the Conway Institute!

Tue, 18 Oct 2022 00:00:00 +0000

After some happy and productive years at Systems Biology Ireland we have moved to the adjacent Conway Institute of Biomolecular and Biomedical Research

The Conway is an interdisciplinary research institute with members from multiple Schools in UCD, we’re the first from the School of Computer Science. The institute houses much of UCD’s ‘omics’ facilities, including the Genomics Core and Proteomics Core along with many research groups doing basic and applied biology research. It’s a nice place to be and we have room for more members if you’re interested in joining.

Narod receiving her commended poster award from Prof. Helen Roche, director of the Conway

We joined in time for this year’s Conway Festival of Research where Anjan and Narod both presented posters that received commendations and Thomas presented at the moderated poster session.

Thomas presenting his poster at the moderated poster session.

Experimental reproducibility limits the correlation between mRNA and protein abundances in tumour proteomic profiles

Thu, 08 Sep 2022 00:00:00 +0000

Comprehensive prediction of robust synthetic lethality between paralog pairs in cancer cell lines

Wed, 15 Dec 2021 00:00:00 +0000

Proteomic signatures of radioresistance: Alteration of inflammation, angiogenesis and metabolism-related factors in radioresistant oesophageal adenocarcinoma

Sun, 18 Apr 2021 00:00:00 +0000

Integrative analysis of large-scale loss-of-function screens identifies robust cancer-associated genetic interactions

Thu, 28 May 2020 00:00:00 +0000

Paralog buffering contributes to the variable essentiality of genes in cancer cell lines

Tue, 01 Jan 2019 00:00:00 +0000

Synthetic Lethality and Cancer – Penetrance as the Major Barrier

Sat, 15 Sep 2018 00:00:00 +0000

Identifying genetic dependencies in cancer by analyzing siRNA screens in tumor cell line panels

Mon, 01 Jan 2018 00:00:00 +0000

A Compendium of Co-regulated Protein Complexes in Breast Cancer Reveals Collateral Loss Events

Wed, 25 Oct 2017 00:00:00 +0000

CancerGD: a resource for identifying and interpreting genetic dependencies in cancer

Wed, 26 Jul 2017 00:00:00 +0000

Genetic Interaction Score (S-Score) Calculation, Clustering, and Visualization of Genetic Interaction Profiles for Yeast

Fri, 21 Jul 2017 00:00:00 +0000

High-Throughput Quantitative Genetic Interaction Mapping in the Fission Yeast Schizosaccharomyces pombe

Fri, 21 Jul 2017 00:00:00 +0000

The histone variant H2A. Z promotes splicing of weak introns

Sat, 01 Apr 2017 00:00:00 +0000

Genome-wide epistatic expression quantitative trait loci discovery in four human tissues reveals the importance of local chromosomal interactions governing gene expression

Sat, 21 Feb 2015 00:00:00 +0000

DAISY: picking synthetic lethals from cancer genomes

Mon, 08 Sep 2014 00:00:00 +0000

High-resolution network biology: connecting sequence with function

Thu, 07 Nov 2013 00:00:00 +0000

All or nothing: protein complexes flip essentiality between distantly related eukaryotes

Wed, 08 May 2013 00:00:00 +0000

The yeast SR-like protein Npl3 links chromatin modification to mRNA processing

Thu, 29 Nov 2012 00:00:00 +0000

Hierarchical Modularity and the Evolution of Genetic Interactomes across Species

Fri, 08 Jun 2012 00:00:00 +0000

Comparative interaction networks: bridging genotype to phenotype

Sun, 01 Jan 2012 00:00:00 +0000

Key Functional Regions in the Histone Variant H2A.Z C-Terminal Docking Domain

Thu, 25 Aug 2011 00:00:00 +0000

Improved functional overview of protein complexes using inferred epistatic relationships

Mon, 23 May 2011 00:00:00 +0000

Imputing and predicting quantitative genetic interactions in epistatic MAPs

Sat, 01 Jan 2011 00:00:00 +0000

Missing value imputation for epistatic MAPs

Tue, 20 Apr 2010 00:00:00 +0000