Functional data analysis approaches for genotype-phenotype association studies from next-generation sequencing

Next-generation DNA sequencing platforms can effectively detect the entire spectrum of genomic variation and is emerging to be a major tool for systematic exploration of the universe of variants and interactions in the entire genome. However, the data produced by next-generation sequencing technologies will suffer from three basic problems: sequence errors, assembly errors, and missing data. Current statistical methods for genetic analysis are well suited for detecting the association of common variants, but are less suitable to rare variants. This raises great challenge for sequence-based genetic studies of complex diseases.^ This research dissertation utilized genome continuum model as a general principle, and stochastic calculus and functional data analysis as tools for developing novel and powerful statistical methods for next generation of association studies of both qualitative and quantitative traits in the context of sequencing data, which finally lead to shifting the paradigm of association analysis from the current locus-by-locus analysis to collectively analyzing genome regions.^ In this project, the functional principal component (FPC) methods coupled with high-dimensional data reduction techniques will be used to develop novel and powerful methods for testing the associations of the entire spectrum of genetic variation within a segment of genome or a gene regardless of whether the variants are common or rare.^ The classical quantitative genetics suffer from high type I error rates and low power for rare variants. To overcome these limitations for resequencing data, this project used functional linear models with scalar response to develop statistics for identifying quantitative trait loci (QTLs) for both common and rare variants. To illustrate their applications, the functional linear models were applied to five quantitative traits in Framingham heart studies. ^ This project proposed a novel concept of gene-gene co-association in which a gene or a genomic region is taken as a unit of association analysis and used stochastic calculus to develop a unified framework for testing the association of multiple genes or genomic regions for both common and rare alleles. The proposed methods were applied to gene-gene co-association analysis of psoriasis in two independent GWAS datasets which led to discovery of networks significantly associated with psoriasis.^

INDUCED-PLURIPOTENT STEM-DERIVED NEURONAL PROGENITOR CELLS AS A NOVEL TREATMENT FOR NEURODEGENERATIVE DISEASES

Cellular therapies, as neuronal progenitor (NP) cells grafting, are promising therapies for patients affected with neurodegenerative diseases like Creutzfeldt-Jakob Disease (CJD). At this time there is no effective treatment or cure for CJD. The disease is inevitably fatal and affected people usually die within months of the appearance of the first clinical symptoms. Compelling evidence indicate that the hallmark event in the disease is the conversion of the normal prion protein (termed PrPC) into the disease-associated, misfolded form (called PrPSc). Thus, a reasonable therapeutic target would be to prevent PrP misfolding and prion replication. This strategy has been applied with poor results since at the time of clinical intervention substantial brain damage has been done. It seems that a more effective treatment aimed at patients with established symptoms of CJD would need to stop further brain degeneration or even recover some of the previously lost brain tissue. The most promising possibility to recover brain tissue is the use of NPs that have the potential to replenish the nerve cells lost during the early stages of the disease. Advanced cellular therapies, beside their potential for cell replacement, might be used as biomaterials for drug delivery in order to stimulate cell survival or the resolution the disease. Also, implanted cells can be genetically manipulated to correct abnormalities causing disease or to make them more resistant to the toxic microenvironments present in damaged tissue. In recent years cell engineering has been within the scope of the scientific and general community after the development of technologies able to “de-differentiate” somatic cells into induced-pluripotent stem (IPS) cells. This new tool permits the use of easy-to-reach cells like skin or blood cells as a primary material to obtain embryonic stem-like cells for cellular therapies, evading all ethical issues regarding the use of human embryos as a source of embryonic stem cells. The complete work proposes to implant IPS-derived NP cells into the brain of prion-infected animals to evaluate their therapeutic potential. Since it is well known that the expression of prion protein in the cell membrane is necessary for PrPSc mediated toxicity, we also want to determine if NPs lacking the prion protein have better survival rates once implanted into sick animals. The main objective of this work is to develop implantable neural precursor from IPS coming from animals lacking the prion protein. Specific aim 1: To develop and characterize cellular cultures of IPS cells from prp-/- mice. Fibroblasts from prp-/- animals will be reprogrammed using the four Yamanaka factors. IPS colonies will be selected and characterized by immunohistochemistry for markers of pluripotency. Their developmental capabilities will be evaluated by teratoma and embryoid body formation assays. Specific aim 2: To differentiate IPS cells to a neuronal lineage. IPS cells will be differentiated to a NP stage by the use of defined media culture conditions. NP cells will be characterized by their immunohistochemical profile as well as by their ability to differentiate into neuronal cells. Specific aim 3: Cellular labeling of neuronal progenitors cells for in vitro traceability. In order to track the cells once implanted in the host brain, they will be tagged with different methods such as lipophilic fluorescent tracers and transduction with GFP protein expression.

MicroRNA Regulation of Prostate Cancer Stem/Progenitor Cells and Prostate Cancer Development

Most human tumors contain a population of cells with stem cell properties, called cancer stem cells (CSCs), which are believed to be responsible for tumor establishment, metastasis, and resistance to clinical therapy. It’s crucial to understand the regulatory mechanisms unique to CSCs, so that we may design CSC-specific therapeutics. Recent discoveries of microRNA (miRNA) have provided a new avenue in understanding the regulatory mechanisms of cancer. However, how miRNAs may regulate CSCs is still poorly understood. Here, we present miRNA expression profiling in six populations of prostate cancer (PCa) stem/progenitor cells that possess distinct tumorigenic properties. Six miRNAs were identified to be commonly and differentially expressed, namely, four miRNAs (miR-34a, let-7b, miR-106a and miR-141) were under-expressed, and two miRNAs (miR-301 and miR-452) were over-expressed in the tumorigenic subsets compared to the corresponding marker-negative subpopulations. Among them, the expression patterns of miR-34, let-7b, miR-141 and miR-301 were further confirmed in the CD44+ human primary prostate cancer (HPCa) samples. We then showed that miR-34a functioned as a critical negative regulator in prostate CSCs and PCa development and metastasis. Over-expression of miR-34a in either bulk or CD44+ PCa cells significantly suppressed clonal expansion, tumor development and metastasis. Systemic delivery of miR-34a in tumor-bearing mice demonstrated a potent therapeutic effect again tumor progression and metastasis, leading to extended animal survival. Of great interest, we identified CD44 itself as a direct and relevant downstream target of miR-34a in mediating its tumor-inhibitory effects. Like miR-34a, let-7 manifests similar tumor suppressive effects in PCa cells. In addition, we observed differential mechanisms between let-7 and miR-34a on cell cycle, with miR-34a mainly inducing G1 cell-cycle arrest followed by cell senescence and let-7 inducing G2/M arrest. MiR-301, on the other hand, exerted a cell type dependent effect in regulating prostate CSC properties and PCa development. In summary, our work reveals that the prostate CSC populations display unique miRNA expression signatures and different miRNAs distinctively and coordinately regulate various aspects of CSC properties. Altogether, our results lay a scientific foundation for developing miRNA-based anti-cancer therapy.

EFFECTS OF THE ACTA2 R258C MUTATION ON VASCULAR SMOOTH MUSCLE CELL PHENOTYPE AND PROPERTIES

Thoracic Aortic Aneurysms and Dissections (TAAD) are the fifteenth leading cause of death in the United States. About 15% of TAAD patients have family history of the disease. The most commonly mutated gene in these families is ACTA2, encoding smooth muscle-specific α-actin. ACTA2 missense mutations predispose individuals both to TAAD and to vascular occlusive disease of small, muscular arteries. Mice carrying an Acta2 R258C mutant transgene with a wildtype Acta2 promoter were generated and bred with Acta2-/- mice to decrease the wildtype: mutant Acta2 ratio. Acta2+/+ R258C TGmice have decreased aortic contractility without aortic disease. Acta2+/- R258C TG mice, however, have significant aortic dilatations by 12 weeks of age and a hyperproliferative response to injury. We characterized smooth muscle cells (SMCs) from bothmouse models under the hypothesis that mutant α-actin has a dominant negative effect, leading to impaired contractile filament formation/stability, improper focal adhesion maturation and increased proliferation. Explanted aortic SMCs from Acta2+/+ R258C TG mice are differentiated - they form intact filaments, express higher levels of contractile markers compared to wildtype SMCs and have predominantly nuclear Myocardin-Related Transcription Factor A (MRTF-A) localization. However, ultracentrifugation assays showed large unpolymerized actin fractions, suggesting that the filaments are brittle. In contrast, Acta2+/- R258C TG SMCs are less well-differentiated, with pools of unpolymerized actin, more cytoplasmic MRTF-A and decreased contractile protein expression compared to wildtype cells. Ultracentrifugation assays after treating Acta2+/- R258C TGSMCs with phalloidin showed actin filament fractions, indicating that mutant α-actin can polymerize into filaments. Both Acta2+/+ R258C TGand Acta2+/- R258C TGSMCs have larger and more peripheral focal adhesions compared to wildtype SMCs. Rac1 was more activated in Acta2+/+ R258C TGSMCs; both Rac1 and RhoA were less activated in Acta2+/- R258C TG SMCs, and FAK was more activated in both transgenic SMC lines compared to wildtype. Proliferation in both cell lines was significantly increased compared to wildtype cells and could be partially attenuated by inhibition of FAK or PDGFRβ. These data support a dominant negative effect of the Acta2 R258C mutation on the SMC phenotype, with increasing phenotypic severity when wildtype: mutant α-actin levels are decreased.

DETERMINING THE GENOTYPE-PHENOTYPE CONNECTION IN SYNTHETIC INDUCIBLE GENE EXPRESSION SYSTEMS

Introduction Gene expression is an important process whereby the genotype controls an individual cell’s phenotype. However, even genetically identical cells display a variety of phenotypes, which may be attributed to differences in their environment. Yet, even after controlling for these two factors, individual phenotypes still diverge due to noisy gene expression. Synthetic gene expression systems allow investigators to isolate, control, and measure the effects of noise on cell phenotypes. I used mathematical and computational methods to design, study, and predict the behavior of synthetic gene expression systems in S. cerevisiae, which were affected by noise. Methods I created probabilistic biochemical reaction models from known behaviors of the tetR and rtTA genes, gene products, and their gene architectures. I then simplified these models to account for essential behaviors of gene expression systems. Finally, I used these models to predict behaviors of modified gene expression systems, which were experimentally verified. Results Cell growth, which is often ignored when formulating chemical kinetics models, was essential for understanding gene expression behavior. Models incorporating growth effects were used to explain unexpected reductions in gene expression noise, design a set of gene expression systems with “linear” dose-responses, and quantify the speed with which cells explored their fitness landscapes due to noisy gene expression. Conclusions Models incorporating noisy gene expression and cell division were necessary to design, understand, and predict the behaviors of synthetic gene expression systems. The methods and models developed here will allow investigators to more efficiently design new gene expression systems, and infer gene expression properties of TetR based systems.

Galectin-3 Enhances the Malignant Melanoma Phenotype by Regulating Autotaxin

In melanoma patient specimens and cell lines, the over expression of galectin-3 is associated with disease progression and metastatic potential. Herein, we have sought out to determine whether galectin-3 affects the malignant melanoma phenotype by regulating downstream target genes. To that end, galectin-3 was stably silenced by utilizing the lentivirus-incorporated small hairpin RNA in two metastatic melanoma cell lines, WM2664 and A375SM, and subjected to gene expression microarray analysis. We identified and validated the lysophospholipase D enzyme, autotaxin, a promoter of migration, invasion, and tumorigenesis, to be down regulated after silencing galectin-3. Silencing galectin-3 significantly reduced the promoter activity of autotaxin. Interestingly, we also found the transcription factor NFAT1 to have reduced protein expression after silencing galectin-3. Electrophoretic mobility shift assays from previous reports have shown that NFAT1 binds to the autotaxin promoter in two locations. ChIP analysis was performed, and we observed a complete loss of bound NFAT1 to the autotaxin promoter after silencing galectin-3 in melanoma cells. Mutation of the NFAT1 binding sites at either location reduces autotaxin promoter activity. Silencing NFAT1 reduces autotaxin expression while over expressing NFAT1 in NFAT1 negative SB-2 melanoma cells induces autotaxin expression. These data suggest that galectin-3 silencing reduces autotaxin transcription by reducing the amount of NFAT1 protein expression. Rescue of galectin-3 rescues both NFAT1 and autotaxin. We also show that the re-expression of autotaxin in galectin-3 shRNA melanoma cells rescues the angiogenic phenotype in vivo. Furthermore, we identify NFAT1 as a potent inducer of tumor growth and experimental lung metastasis. Our data elucidate a previously unidentified mechanism by which galectin-3 regulates autotaxin and assign a novel role for NFAT1 during melanoma progression.

RNA-SEQUENCING APPLICATIONS: GENE EXPRESSION QUANTIFICATION AND METHYLATOR PHENOTYPE IDENTIFICATION

My dissertation focuses on two aspects of RNA sequencing technology. The first is the methodology for modeling the overdispersion inherent in RNA-seq data for differential expression analysis. This aspect is addressed in three sections. The second aspect is the application of RNA-seq data to identify the CpG island methylator phenotype (CIMP) by integrating datasets of mRNA expression level and DNA methylation status. Section 1: The cost of DNA sequencing has reduced dramatically in the past decade. Consequently, genomic research increasingly depends on sequencing technology. However it remains elusive how the sequencing capacity influences the accuracy of mRNA expression measurement. We observe that accuracy improves along with the increasing sequencing depth. To model the overdispersion, we use the beta-binomial distribution with a new parameter indicating the dependency between overdispersion and sequencing depth. Our modified beta-binomial model performs better than the binomial or the pure beta-binomial model with a lower false discovery rate. Section 2: Although a number of methods have been proposed in order to accurately analyze differential RNA expression on the gene level, modeling on the base pair level is required. Here, we find that the overdispersion rate decreases as the sequencing depth increases on the base pair level. Also, we propose four models and compare them with each other. As expected, our beta binomial model with a dynamic overdispersion rate is shown to be superior. Section 3: We investigate biases in RNA-seq by exploring the measurement of the external control, spike-in RNA. This study is based on two datasets with spike-in controls obtained from a recent study. We observe an undiscovered bias in the measurement of the spike-in transcripts that arises from the influence of the sample transcripts in RNA-seq. Also, we find that this influence is related to the local sequence of the random hexamer that is used in priming. We suggest a model of the inequality between samples and to correct this type of bias. Section 4: The expression of a gene can be turned off when its promoter is highly methylated. Several studies have reported that a clear threshold effect exists in gene silencing that is mediated by DNA methylation. It is reasonable to assume the thresholds are specific for each gene. It is also intriguing to investigate genes that are largely controlled by DNA methylation. These genes are called “L-shaped” genes. We develop a method to determine the DNA methylation threshold and identify a new CIMP of BRCA. In conclusion, we provide a detailed understanding of the relationship between the overdispersion rate and sequencing depth. And we reveal a new bias in RNA-seq and provide a detailed understanding of the relationship between this new bias and the local sequence. Also we develop a powerful method to dichotomize methylation status and consequently we identify a new CIMP of breast cancer with a distinct classification of molecular characteristics and clinical features.

Altered tumorigenic phenotype in mice with a hypomorphic p53 mutant allele

Mutations in the p53 tumor suppressor gene are found in over 50% of human tumors and in the germline of Li-Fraumeni syndrome families. About 80% of these mutations are missense in nature. In order to study how p53 missense mutations affect tumorigenesis in vivo, we focused on the murine p53 arg-to-his mutation at amino acid 172, which corresponds to the human hot spot mutation at amino acid 175. The double replacement procedure was employed to introduce the p53 R172H mutation into the p53 locus of ES cells and mice were generated. An additional 1bp deletion in the intron 2 splice acceptor site was detected in the same allele in mice. We named this allele p53R172HΔg. This allele makes a small amount of full length p53 mutant protein. ^ Spontaneous tumor formation and survival were studied in these mice. Mice heterozygous for the p53R172HΔg allele showed 50% survival at 17 months of age, similar to the p53+/− mice. Moreover, the p53R172HΔg/+ mice showed a distinct tumor spectrum: 55% sarcomas, including osteosarcoms, fibrosarcomas and angiosarcomas; 27% carcinomas, including lung adenocarcinomas, squamous cell carcinomas, hepatocellular carcinomas and islet cell carcinomas; and 18% lymphomas. Compared to the p53+/− mice, there was a clear increase in the frequency of carcinoma development and a decrease in lymphoma incidence. Among the sarcomas that developed, fibrosarcomas in the skin were also more frequently observed. More importantly, osteosarcomas and carinomas that developed in the p53R172HΔg/+ mice metastasized at very high frequency (64% and 67%, respectively) compared with less than 10% in the p53+/− mice. The metastatic lesions were usually found in lung and liver, and less frequently in other tissues. The altered tumor spectrum in the mice and increased metastatic potential of the tumors suggested that the p53R172H mutation represents a gain-of-function. ^ Mouse embryonic fibroblasts (MEFs) from the mice homozygous and heterozygous for the p53R172HΔg allele were studied for growth characteristics, immortalization potential and genomic instability. All of the p53R172HΔg /+ MEF lines are immortalized under a 3T3 protocol while under the same protocol p53+/− MEFs are not immortalized. Karyotype analysis showed a persistent appearance of chromosome end-to-end fusion in the MEFs both homozygous and heterozygous for the p53R172HΔg allele. These observations suggest that increased genomic instability in the cells may cause the altered tumor phenotypes. ^

Genome -wide and locus -specific instability

The discovery of expanded simple repeated sequences causing or associated with human disease has lead to a new area of research involved in the elucidation of how the expanded repeat causes disease and how the repeat becomes unstable. ^ To study the genetic basis of the (CTG)n repeat instability in the DMPK gene in myotonic dystrophy (DM1) patients, somatic cell hybrids were constructed between the lymphocytes of DM1 patients and a variety of Chinese hamster ovary (CHO) cell DNA repair gene deficient mutants. By using small pool PCR (SP-PCR), the instability of the (CTG)n can be quantitated for both the frequency and sizes of length change mutations. ^ Additional SP-PCR analysis on 2/11 subclones generated from this original hybrid showed a marked increase in large repeat deletions, ∼50%. A bimodal distribution of repeats was seen around the progenitor allele and at a large deleted product (within the normal range) with no intermediate products present. ^ To determine if the repair capacity of the CHO cell led to a mutator phenotype in the hamster and hybrid clones, SP-PCR was also done on 3 hamster microsatellites in a variety of hamster cell backgrounds. No variant alleles were seen in over 2500 genome equivalents screened. ^ Human-hamster hybrids have long been shown to be chromosomally unstable, yet information about the stability of repeated sequences was not known. To test if repeat instability was associated with either intact or non-intact human chromosomes, more than 300 microsatellite repeats on 13 human chromosomes (intact and non-intact) were analyzed in eight hybrid cells. No variants were seen between the hybrid and patient alleles in the hybrids. ^ To identify whether DM1 patients have a previously undetected level of genome wide instability or if the instability is truly locus specific, SP-PCR was done on 6 human microsatellites within the patient used to make the hybrid cells. No variants were seen in over 1000 genomes screened. ^ These studies show that the somatic cell hybrid approach is a genetically stable system that allows for the determination of factors that could lead to changes in microsatellite instability. It also shows that there is something inherent about the DM1 expanded (CTG)n repeat that it is solely targeted by, as of yet, and unknown mechanism that causes the repeat to be unstable. (Abstract shortened by UMI.)^

Loss of transformed phenotype in cancer cells by overexpression of the uteroglobin gene

Uteroglobin (UG) is a multifunctional, secreted protein that has receptor-mediated functions. The human UG (hUG) gene is mapped to chromosome 11q12.2–13.1, a region frequently rearranged or deleted in many cancers. Although high levels of hUG expression are characteristic of the mucosal epithelia of many organs, hUG expression is either drastically reduced or totally absent in adenocarcinomas and in viral-transformed epithelial cells derived from the same organs. In agreement with these findings, in an ongoing study to evaluate the effects of aging on UG-knockout mice, 16/16 animals developed malignant tumors, whereas the wild-type littermates (n = 25) remained apparently healthy even after 1½ years. In the present investigation, we sought to determine the effects of induced-expression of hUG in human cancer cells by transfecting several cell lines derived from adenocarcinomas of various organs with an hUG-cDNA construct. We demonstrate that induced hUG expression reverses at least two of the most important characteristics of the transformed phenotype (i.e., anchorage-independent growth on soft agar and extracellular matrix invasion) of only those cancer cells that also express the hUG receptor. Similarly, treatment of the nontransfected, receptor-positive adenocarcinoma cells with purified recombinant hUG yielded identical results. Taken together, these data define receptor-mediated, autocrine and paracrine pathways through which hUG reverses the transformed phenotype of cancer cells and consequently, may have tumor suppressor-like effects.

Adult brain retains the potential to generate oligodendroglial progenitors with extensive myelination capacity

Remyelination of focal areas of the central nervous system (CNS) in animals can be achieved by transplantation of glial cells, yet the source of these cells in humans to similarly treat myelin disorders is limited at present to fetal tissue. Multipotent precursor cells are present in the CNS of adult as well as embryonic and neonatal animals and can differentiate into lineage-restricted progenitors such as oligodendroglial progenitors (OPs). The OPs present in adults have a different phenotype from those seen in earlier life, and their potential role in CNS repair remains unknown. To gain insights into the potential to manipulate the myelinating capacity of these precursor and/or progenitor cells, we generated a homogenous culture of OPs from neural precursor cells isolated from adult rat subependymal tissues. Phenotypic characterization indicated that these OPs resembled neonatal rather than adult OPs and produced robust myelin after transplantation. The ability to generate such cells from the adult brain therefore opens an avenue to explore the potential of these cells for repairing myelin disorders in adulthood.

A phenotype-based screen for embryonic lethal mutations in the mouse

The genetic pathways that control development of the early mammalian embryo have remained poorly understood, in part because the systematic mutant screens that have been so successful in the identification of genes and pathways that direct embryonic development in Drosophila, Caenorhabditis elegans, and zebrafish have not been applied to mammalian embryogenesis. Here we demonstrate that chemical mutagenesis with ethylnitrosourea can be combined with the resources of mouse genomics to identify new genes that are essential for mammalian embryogenesis. A pilot screen for abnormal morphological phenotypes of midgestation embryos identified five mutant lines; the phenotypes of four of the lines are caused by recessive traits that map to single regions of the genome. Three mutant lines display defects in neural tube closure: one is caused by an allele of the open brain (opb) locus, one defines a previously unknown locus, and one has a complex genetic basis. Two mutations produce novel early phenotypes and map to regions of the genome not previously implicated in embryonic patterning.

Neuropeptides, by direct interaction with T cells, induce cytokine secretion and break the commitment to a distinct T helper phenotype

Searching for nervous system candidates that could directly induce T cell cytokine secretion, I tested four neuropeptides (NPs): somatostatin, calcitonin gene-related peptide, neuropeptide Y, and substance P. Comparing neuropeptide-driven versus classical antigen-driven cytokine secretion from T helper cells Th0, Th1, and Th2 autoimmune-related T cell populations, I show that the tested NPs, in the absence of any additional factors, directly induce a marked secretion of cytokines [interleukin 2 (IL-2), interferon-γ, IL-4, and IL-10) from T cells. Furthermore, NPs drive distinct Th1 and Th2 populations to a “forbidden” cytokine secretion: secretion of Th2 cytokines from a Th1 T cell line and vice versa. Such a phenomenon cannot be induced by classical antigenic stimulation. My study suggests that the nervous system, through NPs interacting with their specific T cell-expressed receptors, can lead to the secretion of both typical and atypical cytokines, to the breakdown of the commitment to a distinct Th phenotype, and a potentially altered function and destiny of T cells in vivo.

Overexpression of DNA polymerase β in cell results in a mutator phenotype and a decreased sensitivity to anticancer drugs

DNA polymerase β (pol β) is the most error prone of all known eukaryotic DNA polymerases tested in vitro. Here, we show that cells overexpressing pol β cDNA have acquired a spontaneous mutator phenotype. By measuring the appearance of mutational events using three independent assays, we found that genetic instability increased in the cell lines that overexpressed pol β. In addition, these cells displayed a decreased sensitivity to cancer chemotherapeutic, bifunctional, DNA-damaging agents such as cisplatin, melphalan, and mechlorethamine, resulting in enhanced mutagenesis compared with control cells. By using cell-free extracts and modified DNA substrates, we present data in support of error-prone translesion replication as one of the key determinants of tolerance phenotype. These results have implications for the potential role of pol β overexpression in cancer predisposition and tumor progression during chemotherapy.

Genetic and root phenotype diversity in Sri Lankan rice landraces may be related to drought resistance

Acknowledgements Mayuri Munasinghe was supported by a Commonwealth Scholarship (ref no. LKCS-2009-384). The development and use of the SNP chip was funded by a BBSRC grant BB/J003336/1. The authors thank Owen Price (University of Wollongong, Australia) for producing the coloured province map of Sri Lanka, Gareth Norton (Aberdeen) for merging the RDP1 SNP data with the Sri Lankan data and Tony Travis (Aberdeen) for help with PCA.