Gary Hardiman Queens University Belfast
Philippe Lemey K.U. Leuven
Clare West BenevolentAI
Orla O'Sullivan APC Microbiome
Niamh Mullins Icahn School of Medicine, Mount Sinai
Dan Bradley Trinity College Dublin
Single-cell characterisation of the hematopoietic bone marrow niche in health and disease. Sarah Ennis \ Assessing the Role of Genomics Data in Stratifying Patients within Predictive Models for Breast Cancer Survival Outcome. Lydia King \ Genetically Predicted Circulating Concentrations of Micronutrients and COVID-19 Susceptibility and Severity: A Mendelian Randomization Study. Neil Daniel \ Exploring the role of microRNA-associated variants in Multiple Sclerosis. Ifeolutembi Fashina \ Burden for autoimmune diseases as quantified by Polygenic Risk Scoring (PRS) as a predictor for severe cutaneous adverse drug reactions attributable to aromatic anti-epileptic drugs. Isabelle Boothman \ Polygenic burden for intracranial aneurysm and hypertension in deceased kidney donors who died of intracranial haemorrhage. Kane Collins \ Functional and population genomics of admixed trypanotolerant African cattle breeds. Gillian P. McHugo \ The Search for Schizosaccharomyces fission yeasts in environmental meta-transcriptomes. Rasha Shraim \ PMI - A software tool for entity based organisational modelling and biological systems analysis. Ruan O'Tiarnaigh \ Rectal cancer-specific subtyping using transcriptomics datasets. Batuhan Kisakol Induced torpor may offer radio-protective effects through the induction of the hypoxia response in zebrafish. Thomas Cahill \ Selective and Demographic Factors associated with Neanderthal Introgressed Tract Lengths in Europeans and East Asians. Elle Loughran
The continued efficacy of antibiotics is currently uncertain due to the global dissemination of antibiotic-resistance determinants. Without urgent action, it is projected that deaths attributable to resistant infections will reach 10 million per year by 2050. Among the main antibiotic-resistant bacteria, Klebsiella pneumoniae is specially determinant in human health. The main purpose of this research is to discover novel groups of predicted antibiotic compounds by applying a chemoinformatics algorithm suited to machine learning targeting Klebsiella pneumoniae. Based on the Minimum Inhibitory Concentrations available in the ChEMBL database, 12250 compounds have been classified into active, medium and inactive compounds (1510, 1487 and 9254 compounds, respectively). For each one, the molecule's representation has been encoded into a set of 80 selected molecular descriptors that has been applied to different neural networks to predict the antibiotic activity of the compounds. To train the different networks, different numbers of layers, activation functions (ReLU, sigmoid and tanh) and optimizers (ADAM and SGD) have been applied. The results show that models trained with the sigmoid function are more unstable, those with the tanh function are slower to train than the ones with the ReLU function, and those with higher number of layers and units per layer are slower to train but have better predictive power. In the same way, the ADAM optimizer is faster than the SGD and seems to have more predictive power. A selection of those models show a high predictive power, with AUC values greater than 95% in some cases, when testing them.
Identifying risk factors for precancerous colorectal adenoma (CRA) development is essential for improving colorectal cancer (CRC) screening and prevention. Some adenomas can present a mucous cap that may offer a niche for bacterial growth that could contribute to dysplastic progression. Detailed investigation is required to define microbiota patterns during the CRA to CRC transition. We assessed the adenoma microbial profiles in colorectal neoplasia (n=14, grouping samples obtained from tissue and colonic brushes) vs control (n=9, normal mucosa at a distinct site). Furthermore, we also investigated taxonomic differences between the mucosal surface of the adenoma (n=9) and the tissue (n=5). Sequencing of V3-V4 16S rRNA region was performed to determine the bacterial composition. Shannon index (H) was used to calculate the alpha diversity within the samples. Although the two pathology groups did not show substantial differences, disease samples from colonic brushes of the surface of the adenoma presented a lower H compared to the fresh frozen tissues (Hmean_brushes_disease=1.27, Hmean__tissue_disease=2.47). Differential abundance analysis showed that Bacteroides intestinalis and Bacteroides fragilis were depleted in disease compared to the controls (Log2 fold changes= -1.86, = -2.49, respectively) while Prevotella (species na), Bacteroides vulgatus, Bacteroides uniformis and Alistipes massilensis were enriched in colorectal neoplasia samples (Log2 fold changes= 2.51, =2.06, =1.73, =1.39, = 1.71, =1.35, = 1.41, respectively). Eighteen species were statistically different between the mucosal surface of the adenoma and the tissue. All of them were depleted in tissue compared to the adenoma surface. Bacteroides thetaiotaomicron, Bacteroides vulgatus, Bacteroides fragilis and Alistipes putredinis presented the highest log2 fold changes (Log2 fold changes= -5.88, = -3.70, =-4.84, respectively). These results indicate that the gut microbiome present on the surface of the adenoma and the microbiome in the tissue may be compositionally different. In future, a functional characterization of the microbiome combining metatranscriptomic and metagenomic in colorectal adenomas will be performed to clarify whether adenomas progression is influenced by microbial dysbiosis.
Gene loss is a common feature of tumour cells. It is unclear how cellular networks in tumour cells rewire to accommodate such gene loss. Paralogs (duplicate genes) may provide a means of robustness in tumour cells. Paralogs can have compensatory or 'collateral loss' relationships with each other, where one paralog is either upregulated (compensation) or downregulated (collateral loss) when the other is lost. Further, these relationships might only be observable at the protein level, not the gene/mRNA level. Here, we used the Cancer Cell Line Encyclopaedia proteomic dataset, containing expression data for 12755 proteins in 375 cancer cell lines, to identify such relationships between paralogs. We reprocessed this dataset to obtain protein quantifications using only unique peptides, i.e. peptides that only map to one protein. This was necessary since paralogs might share several peptides. We then used a regression model to look for cases where, in a paralog pair A1-A2, loss of A2 (based on mutation and transcriptomic data) was associated with a significant (FDR < 0.1) increase or decrease in A1 expression. We found 121 such cases for compensation (A1 increase) and 527 cases for collateral loss (A1 decrease). A Fisher's Exact Test showed that protein complex membership made paralog pairs significantly (p < 0.01) more likely to have a compensatory relationship. Similarly, if A2 loss was associated with an increase in A1 essentiality (based on DepMap genome-wide CRISPR screen data), it was significantly (p < 0.01) more likely that A2 loss would be associated with an increase in A1 expression.
A limited number of Mendelian genetic conditions are diagnosable in the archaeological record. One such condition is Multiple Osteochondromas, a rare autosomal dominant disease resulting in the formation of bony tumours in affected individuals. This has been identified in a handful of individuals from the Middle Bronze Age to the post-medieval period. Two individuals from the Gaelic Medieval cemetery of Ballyhanna, Co. Donegal (dated AD 690-885 and AD 1030-1220) showed evidence of bone tumours consistent with this condition. Whole-genome shotgun sequencing revealed a missense mutation in the second exon of the gene EXT1 in the earlier individual, which has been identified previously in present-day patients. The second individual did not carry this mutation, but a different, frameshift mutation in the first exon of the same gene was identified, leading to a premature stop and loss of function. The difference between these individuals is surprising: just 10% of modern cases are suspected to be caused by de novo mutation, and clusters of this disease have been identified in modern isolated populations, suggesting founder effects, which would have been expected in the same cemetery in northwestern Medieval Ireland.
Segmentally duplicated regions have been systematically excluded from genetic association studies due to technical difficulties. These regions are the main source of copy number variation in the genome and are implicated in the emergence of novel genes underpinning human brain evolution and as risk factors in neuropathologies. The human Pregnancy-Specific Glycoprotein (PSG) locus on 19q32 comprises a set of segmental duplications arising from the evolution of 11 protein-coding genes with high sequence identity. PSG genes are expressed in the placenta and gut and their main functions relate to immunomodulation. In addition, the PSG locus contains several non-coding genes which are predominantly expressed in the brain. PSG8-AS1 is a great ape-specific lncRNA that we have implicated in oligodendrocyte differentiation and myelination and which could be associated with Multiple Sclerosis. Common deletions of a different region of the PSG locus may be associated with Schizophrenia, based on our preliminary analysis of publicly available datasets. The deleted region contains CEACAMP10, and interrupts PSG4, which are expressed in the brain in some cell types; and an expression quantitative trait locus (eQTL) for PAFAH1B1, a protein involved in neurogenesis.
Gaining insight into the mechanisms of signal transduction networks (STNs) by using critical features from patient-specific mathematical models can improve patient stratification and help to identify potential drug targets. To achieve this, these models should focus on the critical STNs for each cancer, include prognostic genes and proteins, and correctly predict patient-specific differences in STN activity. Focusing on colorectal cancer and the WNT STN, we used mechanism-based machine learning models to identify genes and proteins with significant associations to event-free patient survival and predictive power for explaining patient-specific differences of STN activity. First, we identified the WNT pathway as the most significant pathway associated with event-free survival. Second, we built linear-regression models that incorporated both genes and proteins from established mechanistic models in the literature and novel genes with significant associations to event-free patient survival. Data from The Cancer Genome Atlas and Clinical Proteomic Tumour Analysis Consortium were used, and patient-specific STN activity scores were computed using PROGENy. Three linear regression models were built, based on; (1) the gene-set of a state-of-the-art mechanistic model in the literature, (2) novel genes identified, and (3) novel proteins identified. The novel genes and proteins were genes and proteins of the extant WNT pathway whose expression was significantly associated with event-free survival. The results show that the predictive power of a model that incorporated novel event-free associated genes is better compared to a model focusing on the genes of a current state-of-the-art mechanistic model. Several significant genes that should be integrated into future mechanistic models of the WNT pathway are DVL3, FZD5, RAC1, ROCK2, GSK3B, CTB2, CBT1, and PRKCA. Thus, the study demonstrates that using mechanistic information in combination with machine learning can identify novel features (genes and proteins) that are important for explaining the STN heterogeneity between patients and their association to clinical outcomes.
Long access heroin self-administration significantly alters gut microbiome diversity, structure and composition. James Mackle \ Fusion genes in prostate cancer. A comparison in men of African and European descent. Rebecca Morgan \ Statistical Detection of Meiotic Recombination Events. Catherine Mahoney \ An automated computational pipeline for novel epileptic encephalopathy gene discovery using whole exome sequencing trios. Hamidah Ghani \ The Contribution of Other hits in ND-CNV Carriers to Cognitive Outcomes. Thomas J. Dinneen \ Harnessing the power of whole animal single cell sequencing to understand genome stability maintenance in an immortal animal. Helen Horkan \ Gene regulatory network-inferred driver activity as a predictor of cancer cells sensitivity to gene inhibition. Cosmin Tudose \ Representation of the Protein Coding Information in RNA transcripts with Ribosome Decision Graphs. Jack Tierney \ Machine Learning and Multi-Omics to Predict Relapse in Inflammatory Bowel Disease. Jill O'Sullivan \ More accurate and unbiased associations in genetic association studies. Amanda Forde \ Genome-wide local ancestry and direct evidence for mitonuclear co-1 adaptation in African hybrid cattle populations (Bos taurus/indicus). James A. Ward
Killer plasmids are linear, double-stranded DNA virus-like elements found in several species of yeast. A yeast possessing a killer plasmid secretes toxins, anticodon nucleases (ACNases), which destroy specific tRNAs in non-killer yeasts. These ACNase toxins may be important candidates for anticancer and antimicrobial research and are hypothesized to be responsible for the unusual genetic codes used in some yeast species, where the codon CUG is read by a tRNA-Ser-CAG or a tRNA-Ala-CAG instead of tRNA-Leu-CAG. However, only 4 ACNase toxins are known, sharing ~30% amino acid identity due to rapid evolution, and none have been shown to target tRNA-Leu-CAG. Killer plasmids are rare in the natural environment (found in only 1-2% of yeast strains) and are often missing from genome sequences due to their high coverage and short length.I developed a pipeline to automatically download and assemble raw data from every publicly available budding yeast genome from the NCBI Sequence Repository Archive (SRA). We searched (TBLASTN) these clades for moderately conserved genes found within Killer Plasmids to find candidate toxins in linkage to these genes. Candidates were identified based upon the presence of a secretion signal and residues essential for toxicity. Using this approach, we have identified 29 new candidate toxin genes and 8 novel killer plasmids, with several hits demonstrating plasmid integration into the host nucleus. We will evaluate these candidate ACNases and their tRNA targets in the wet lab in the coming months.
Bovine Respiratory disease (BRD) affects cattle of all ages, causing significant morbidity and mortality in Ireland, and internationally. Bovine Herpes Virus 1 (BoHV-1), a DNA virus belonging to the Alphaherpesvirinae subfamily, is a key virus associated with disease onset. The objective of this study was to examine the response of the host’s bronchial lymph node and cranial lung lobe transcriptome to an experimental challenge with BoHV-1, using RNA-seq. Briefly, calves (Holstein-Friesian) were either administered BoHV-1 inoculate (1x107/ml x 8.5ml) (n=12) or mock challenged with sterile phosphate buffered saline (PBS) (n=6). Clinical signs were recorded daily and calves were euthanized on day 6 post challenge. On euthanisation, lungs were scored for the presence of lesions and tissues collected. RNA was extracted from the bronchial lymph node and cranial lung lobe tissues and sequenced (150bp paired-end). Sequenced reads were aligned to the bovine reference genome (ARS-UCD 1.2) and EdgeR was used for differential expression analysis. A clear separation was seen between the BoHV-1 challenged and control calves using multi-dimensional scaling (MDS). A total of 337 genes were differentially expressed (DE) in the bronchial lymph node between the two treatments. The healthy cranial lobe tissue displayed a total of 334 DE genes between the two treatments, and a total of 67 genes were DE in the lesioned lung lobe. Key KEGG pathways enriched for DE genes include ‘Influenza A, and ‘Herpes Simplex Infection’. Enriched gene ontology terms for the DEG’s, were involved with immune responses. An extensive investigation of the host response to BoHV-1 challenge may lead to the development of new therapeutic targets.
Next generation sequencing studies are dependent on a high-quality reference genome for single nucleotide variant (SNV) calling. Although the two most recent builds of the human genome are widely used (GRCh37 and GRCh38), position information is typically not directly comparable between them. Tools such as liftOver and CrossMap are used to convert data from one build to another. However, the positions of converted SNVs do not always match SNVs derived from aligned data and in some instances, SNVs are known to change chromosome when converted. This inconsistency can be a significant problem for bioinformatics analyses when compiling sequencing resources or comparing results across studies. Here we describe a novel algorithm to identify positions that are unstable when converting between the above reference genome builds. This resulted in 11.3Mb of unstable positions with GRCh37 as the source and 20Mb with GRCh38 as the source. The unstable positions are detected independent of the conversion tools and are determined by the chain files. As a proof of principle, we examined SNVs derived from whole genome sequencing of two publicly available genomes aligned to both builds. Pre-excluding SNVs at these unstable positions prior to conversion results in SNVs that are stable to the conversion process. Variants at unstable positions, although fewer in number, had higher discordance rates with aligned data than variants at stable positions. This work highlights the care that must be taken when converting SNVs between genome builds and provides a simple method for ensuring higher confidence converted data.
Schizophrenia is a psychiatric disorder that affects 1% of adults and is a major global health problem. Altered gene expression in the brain has been associated with neuropsychiatric disorders. Further analysis of gene expression data to isolate and study predicted cell types could uncover further insights into altered gene expression that contributes to schizophrenia etiology. Cellular deconvolution is used to estimate the proportion of cell subtypes present in bulk expression data. Here, we reanalyzed gene expression data from the PsychENCODE consortium (n=558 schizophrenia cases and 1039 controls; cortical samples) by using new single-cell sequencing data that allowed for an improved cellular deconvolution analysis. New single cell data for ~30,000 cells was added to the original dataset to give an increased sample of ~62,000 cells. In comparison to the original analysis, we observed alterations in astrocyte proportions between cases and controls, particularly for two distinct astrocyte populations, only one of which had significantly different proportions between cases and controls (p=<2.2e-16) with a lower proportion of cells inferred in cases. Oligodendrocytes did not have significantly different proportions between cases and controls (p=0.588) as an overall subtype but four distinct subtypes of oligodendrocytes did appear to exhibit differences in proportions (p<0.05). Overall, this analysis leveraged new single-cell data to perform a more detailed cellular deconvolution analysis of data from PsychENCODE and identified a number of distinct cell subtypes that differ in proportion between cases and controls. These can be followed up to uncover new insights into the biological functions that influence schizophrenia.
The Medieval period was a time of increased population mobility and urbanization across Europe. Despite historical accounts for the Medieval period, there is still uncertainty as to the extent to which these changes affected the demography of different regions, including Scotland. The sequencing and analysis of ancient genomes allows us to interrogate Medieval population dynamics and migration, as well as the social structures and kinship dynamics that existed within specific cemetery sites. To this end, we examined burials from an urban graveyard within The Kirk of St.Nicholas in Medieval Aberdeen, sampling evenly from two distinct archaeological phases; phase A (from 12th -15th century) and phase B (15th-18th century). The Kirk of St.Nicholas is a significant cemetery as it was one of the largest burgh churches in Scotland and the primary place of worship in a city with increasing wealth during the Medieval period. We sequenced 30 individuals to a genomic coverage (0.01-3X), allowing us to assess kinship and inbreeding within the site, as well as the population’s broader genetic affinities to other Medieval and present-day populations. Using principal components analysis and formal tests of admixture, we are able to identify non-local individuals, providing a window into migration dynamics at the time. We detected kinship among the burials spanning several centuries suggesting the Kirk cemetery was utilised by an extended familial group with access to prestigious burial.
Previous studies utilising individuals annotated by geographic origin have demonstrated subtle but identifiable genetic structure within the Irish population and have detected admixture signals indicative of gene flow consistent with historical migrations into Ireland. However, these studies were individually limited by sample size, which in turn limited the resolution of their haplotype-based population structure analyses. We therefore set out to assemble a large sample of Irish ancestry references with geographic-origin annotations to expand our understanding of fine-scale genetic population structure and offer preliminary insights into the demographic history of Ireland using British comparisons. Datasets from four studies were combined with 6724 individuals in total - Republic of Ireland (n=3274), Northern Ireland (n=102), England (n=2580), Wales (n=329), Scotland (n=210), Orkney Islands (n=193), and the Isle of Man (n=63). To our knowledge, the combined dataset contains the largest collection of Irish genotype array data with geographical provenance. Leveraging patterns of Identity-By-Descent (IBD) segment sharing, we applied the Leiden algorithm to identify genetic clusters in the dataset. Consistent with previous reports, we observed the population substructures separate along geographic boundaries. Additionally, we calculated levels of Runs-Of-Homozygosity (ROH) and IBD-sharing within these genetic clusters and inferred effective population sizes using IBDNe to reveal preliminary insights into recent demographic history. We intend use the results from these analyses to help disentangle the relative effects population structure and demographic history on the genetic architecture of complex diseases, such as epilepsy or amyotrophic lateral sclerosis (ALS).
Philippe Lemey Prinicipal Investigator, K.U. Leuven
Philippe Lemey's research interests lie in the fields of molecular epidemiology, computational biology and viral evolution. His team studies evolutionary processes that shape viral genetic diversity, spanning from large-scale epidemic processes, such as population growth and spatial dispersal, to small-scale transmission histories and within-host evolutionary processes, including adaptation and recombination. To investigate these different aspects of viral evolution, the team develops and integrates both molecular and computational biology approaches. The team is an important contributor to BEAST ([https://github.com/beast-dev/beast-mcmc](https://github.com/beast-dev/beast-mcmc)), a state-of-the-art Bayesian inference software and has developed phylodynamic models that are widely used for inference of temporal, spatial and adaptation dynamics of virus epidemics.
The human intestinal tract is host to an extensive population of microorganisms working in concert to play a pivotal role in human health. Almost every aspect of modern lifestyles can impact the gut microbiota; recently diet and fitness have been established as important modulators. Previously, we observed that elite athletes have significantly increased gut microbial diversity compared to non-athlete controls. Functional metagenomic analysis revealed this elevated diversity translated into the athletes’ microbiome being primed for energy harvest as well as muscle and tissue repair. However, a subsequent 8-week exercise intervention study failed to reproduce the same high microbial diversity. This lead us to hypothesize that it’s physical fitness, not exercise, that is pivotal to increased microbial diversity. We propose to mine datasets to identify “fitness” associated microbes and metabolites and investigate their implications for human health.
Orla O'Sullivan Research Fellow, APC Microbiome
Orla O'Sullivan completed her PhD in Bioinformatics in 2004 at University College Cork. She is currently a Senior Research Officer (Computational Biologist) at Teagasc and a Principal Investigator at SFI research centres APC Microbiome Ireland and VistaMilk. Her research focuses on the role of the microbiome in diverse environments and the role of exercise and diet on the human gut microbiome.
Human health and disease
Clare West Senior Scientist, BenevolentAI
Clare West is a computational biologist with experience in structural biology and drug discovery. Her postdoctoral research aimed to harness publicly available data to identify and prioritise tractable drug targets within the mechanisms linking ageing and age-related diseases, in collaboration with UK SPINE, the Centre for Medicines Discovery (CMD) at the University of Oxford, and Open Targets at EMBL-EBI. She has a BSc in Biochemistry (University of Nottingham) and a PhD in Protein Informatics (Department of Statistics, University of Oxford) focussing on template-free protein structure prediction.
Gary Hardiman Professor, Queens University Belfast
Prof. Gary Hardiman joined the Faculty of Medicine, Health and Life Sciences and Institute for Global Food Security (IGFS) at Queen’s University Belfast in 2018 as the Chair in Food Systems Biology. His laboratory at QUB works in the field of systems biology the objective of which is the study of biological systems, including genes, RNAs, proteins, metabolites, and cells in a focused manner, and organs, organisms, and populations in a broader context. Current areas of research focus are studying the effects of microplastics on marine and human health; prostate cancer in the context of racial differences and nutritional deficiency; examining the impacts of long-term space travel – specifically the effects of nutrition, torpor, space radiation and microgravity on hepatic and intestinal biology; developing a rat model of opioid abuse; optimizing toolkits for better integration of Omics data sets into genotype-phenotype predictions; and computational analysis to interpret the development of metabolic diseases mediated by non-coding RNAs. Hardiman is co-founder and CTO of the Northern Ireland based precision medicine company Altomics Datamation Ltd. Before moving to Belfast, Hardiman spent his academic career in California and South Carolina, USA. Most recently Hardiman was Scientific Director of the Center for Genomics Medicine Bioinformatics and a full professor in the Departments of Medicine and Public Health Sciences at The Medical University of South Carolina. He was also Head, Laboratory for Marine Systems Biology, Hollings Marine Laboratory, an Adjunct Professor & Visiting Scholar, Grice Marine Laboratory, College of Charleston. He held faculty appointments at the Bioinformatics & Medical Informatics Research Center (BMIRC) and Computational Science Research Center (CSRC), San Diego State University, San Diego and the Department of Medicine at University of California, San Diego. He was the Founding Director of the UCSD Biomedical Genomics Facility (BIOGEM). He served on the advisory boards of several companies including OnRamp Bioinformatics, San Diego, CA; Axikin, La Jolla, CA; Autogenomics, Carlsbad, CA; Molecular Stamping, Trento, Italy. Before returning to academia in 2000, he worked in the biotech industry as a Senior Scientist, at Axys Pharmaceuticals (Quest Diagnostics) in Cambridge, MA; La Jolla, CA & San Francisco, CA. Hardiman received his BSc. Hons. & Ph.D. degrees in Microbiology/Molecular Biology from the National University of Ireland Galway (NUIG) in 1989 and 1993, respectively. He was a University of California/National University of Ireland Education Abroad Scholar and received graduate training at the University of California, Santa Cruz (1992). He completed two post-doctoral research fellowships at DNAX Research Institute (MERCK), Palo Alto, CA (1993-1998) in genomics and bioinformatics. Hardiman serves on the editorial board of the journals ‘Pharmacogenomics’, ‘Expert Review of Molecular Diagnostics’, ACS ES&T Water, Science of the Total Environment (STOTEN), and the MDPI published journals ‘Genes’ and ‘Biotech’ and is an associate editor of ‘Frontiers in Psychiatry -Psychopharmacology’. He has edited three books on genomics technologies, ‘Microarray Methods and Applications’ published by DNA Press, Inc. (2003), ‘Biochips as Pathways to Drug Discovery’ with Dr. Andrew Carmen, published by CRC Press (2007) and ‘Microarray Innovations - Technology & Experimentation’ (2009). He has collaborated with NASA GeneLab, The Gulf of Mexico Research Initiative (GoMRI) on the Deepwater Horizon incident, and the Lipid Maps Consortium.
Dan Bradley Professor and Head of School of Genetics and Microbiology, Trinity College Dublin
His research interests include detection of signatures of selection in human, bovine, salmon and chicken genomes, Irish human population structure, Y chromosome diversity and Irish medieval genealogies, ancient DNA, origins of livestock as discerned using genetic diversity, genetic basis of disease resistance in West African cattle, haplotype decay and the age of cattle hybrid zones, and detection of recombination in mitochondrial genomes.
Niamh Mullins Assistant Professor, Icahn School of Medicine, Mount Sinai
Niamh Mullins is Assistant Professor at the Icahn School of Medicine at Mount Sinai. Her research interests lie in psychiatric genomics, particularly in conducting large-scale genetic studies of bipolar disorder and suicidality. She graduated with a PhD from King's College London in 2017, and now works with the Bipolar Disorder Working Group of the Psychiatric Genomics Consortium and is Co-founder of the International Suicide Genetics Consortium.