PyPop: a mature open-source software pipeline for population genomics.

Python for Population Genomics (PyPop) is a software package that processes genotype and allele data and performs large-scale population genetic analyses on highly polymorphic multi-locus genotype data. In particular, PyPop tests data conformity to Hardy-Weinberg equilibrium expectations, performs Ewens-Watterson tests for selection, estimates haplotype frequencies, measures linkage disequilibrium, and tests significance. Standardized means of performing these tests is key for contemporary studies of evolutionary biology and population genetics, and these tests are central to genetic studies of disease association as well. Here, we present PyPop 1.0.0, a new major release of the package, which implements new features using the more robust infrastructure of GitHub, and is distributed via the industry-standard Python Package Index. New features include implementation of the asymmetric linkage disequilibrium measures and, of particular interest to the immunogenetics research communities, support for modern nomenclature, including colon-delimited allele names, and improvements to meta-analysis features for aggregating outputs for multiple populations. Code available at: https://zenodo.org/records/10080668 and https://github.com/alexlancaster/pypop.

Comparison between qPCR and RNA-seq reveals challenges of quantifying HLA expression.

Human leukocyte antigen (HLA) class I and II loci are essential elements of innate and acquired immunity. Their functions include antigen presentation to T cells leading to cellular and humoral immune responses, and modulation of NK cells. Their exceptional influence on disease outcome has now been made clear by genome-wide association studies. The exons encoding the peptide-binding groove have been the main focus for determining HLA effects on disease susceptibility/pathogenesis. However, HLA expression levels have also been implicated in disease outcome, adding another dimension to the extreme diversity of HLA that impacts variability in immune responses across individuals. To estimate HLA expression, immunogenetic studies traditionally rely on quantitative PCR (qPCR). Adoption of alternative high-throughput technologies such as RNA-seq has been hampered by technical issues due to the extreme polymorphism at HLA genes. Recently, however, multiple bioinformatic methods have been developed to accurately estimate HLA expression from RNA-seq data. This opens an exciting opportunity to quantify HLA expression in large datasets but also brings questions on whether RNA-seq results are comparable to those by qPCR. In this study, we analyze three classes of expression data for HLA class I genes for a matched set of individuals: (a) RNA-seq, (b) qPCR, and (c) cell surface HLA-C expression. We observed a moderate correlation between expression estimates from qPCR and RNA-seq for HLA-A, -B, and -C (0.2 ≤ rho ≤ 0.53). We discuss technical and biological factors which need to be accounted for when comparing quantifications for different molecular phenotypes or using different techniques.

Demographic history and selection at HLA loci in Native Americans.

The American continent was the last to be occupied by modern humans, and native populations bear the marks of recent expansions, bottlenecks, natural selection, and population substructure. Here we investigate how this demographic history has shaped genetic variation at the strongly selected HLA loci. In order to disentangle the relative contributions of selection and demography process, we assembled a dataset with genome-wide microsatellites and HLA-A, -B, -C, and -DRB1 typing data for a set of 424 Native American individuals. We find that demographic history explains a sizeable fraction of HLA variation, both within and among populations. A striking feature of HLA variation in the Americas is the existence of alleles which are present in the continent but either absent or very rare elsewhere in the world. We show that this feature is consistent with demographic history (i.e., the combination of changes in population size associated with bottlenecks and subsequent population expansions). However, signatures of selection at HLA loci are still visible, with significant evidence selection at deeper timescales for most loci and populations, as well as population differentiation at HLA loci exceeding that seen at neutral markers.

Asymmetric linkage disequilibrium: Tools for assessing multiallelic LD.

Standard measures of linkage disequilibrium (LD) provide an incomplete description of the correlation between two loci. Recently, Thomson and Single (2014) described a new asymmetric pair of LD measures (ALD) that give a more complete description of LD. The ALD measures are symmetric and equivalent to the correlation coefficient r when both loci are bi-allelic. When the numbers of alleles at the two loci differ, the ALD measures capture this asymmetry and provide additional detail about the LD structure. In disease association studies the ALD measures are useful for identifying additional disease genes in a genetic region, by conditioning on known effects. In evolutionary genetic studies ALD measures provide insight into selection acting on individual amino acids of specific genes, or other loci in high LD (see Thomson and Single (2014) for these examples). Here we describe new software for computing and visualizing ALD. We demonstrate the utility of this software using haplotype frequency data from the National Marrow Donor Program (NMDP). This enhances our understanding of LD patterns in the NMDP data by quantifying the degree to which LD is asymmetric and also quantifies this effect for individual alleles.

Conditional asymmetric linkage disequilibrium (ALD): extending the biallelic r2 measure.

For multiallelic loci, standard measures of linkage disequilibrium provide an incomplete description of the correlation of variation at two loci, especially when there are different numbers of alleles at the two loci. We have developed a complementary pair of conditional asymmetric linkage disequilibrium (ALD) measures. Since these measures do not assume symmetry, they more accurately describe the correlation between two loci and can identify heterogeneity in genetic variation not captured by other symmetric measures. For biallelic loci the ALD are symmetric and equivalent to the correlation coefficient r. The ALD measures are particularly relevant for disease-association studies to identify cases in which an analysis can be stratified by one of more loci. A stratified analysis can aid in detecting primary disease-predisposing genes and additional disease genes in a genetic region. The ALD measures are also informative for detecting selection acting independently on loci in high linkage disequilibrium or on specific amino acids within genes. For SNP data, the ALD statistics provide a measure of linkage disequilibrium on the same scale for comparisons among SNPs, among SNPs and more polymorphic loci, among haplotype blocks of SNPs, and for fine mapping of disease genes. The ALD measures, combined with haplotype-specific homozygosity, will be increasingly useful as next-generation sequencing methods identify additional allelic variation throughout the genome.

Utilization of neoadjuvant chemotherapy varies in the treatment of women with invasive breast cancer.

BACKGROUND: Treatment with neoadjuvant chemotherapy (NAC) has made it possible for some women to be successfully treated with breast conservation therapy (BCT ) who were initially considered ineligible. Factors related to current practice patterns of NAC use are important to understand particularly as the surgical treatment of invasive breast cancer has changed. The goal of this study was to determine variations in neoadjuvant chemotherapy use in a large multi-center national database of patients with breast cancer. METHODS: We evaluated NAC use in patients with initially operable invasive breast cancer and potential impact on breast conservation rates. Records of 2871 women ages 18-years and older diagnosed with 2907 invasive breast cancers from January 2003 to December 2008 at four institutions across the United States were examined using the Breast Cancer Surgical Outcomes (BRCASO) database. Main outcome measures included NAC use and association with pre-operatively identified clinical factors, surgical approach (partial mastectomy [PM] or total mastectomy [TM]), and BCT failure (initial PM followed by subsequent TM). RESULTS: Overall, NAC utilization was 3.8%l. Factors associated with NAC use included younger age, pre-operatively known positive nodal status, and increasing clinical tumor size. NAC use and BCT failure rates increased with clinical tumor size, and there was significant variation in NAC use across institutions. Initial TM frequency approached initial PM frequency for tumors >30-40 mm; BCT failure rate was 22.7% for tumors >40 mm. Only 2.7% of patients undergoing initial PM and 7.2% undergoing initial TM received NAC. CONCLUSIONS: NAC use in this study was infrequent and varied among institutions. Infrequent NAC use in patients suggests that NAC may be underutilized in eligible patients desiring breast conservation.

Factors associated with the frequency of initial total mastectomy: results of a multi-institutional study.

BACKGROUND: Several previous studies have reported conflicting data on recent trends in use of initial total mastectomy (TM); the factors that contribute to TM variation are not entirely clear. Using a multi-institution database, we analyzed how practice, patient, and tumor characteristics contributed to variation in TM for invasive breast cancer. STUDY DESIGN: We collected detailed clinical and pathologic data about breast cancer diagnosis, initial, and subsequent breast cancer operations performed on all female patients from 4 participating institutions from 2003 to 2008. We limited this analysis to 2,384 incident cases of invasive breast cancer, stages I to III, and excluded patients with clinical indications for mastectomy. Predictors of initial TM were identified with univariate analyses and random effects multivariable logistic regression models. RESULTS: Initial TM was performed on 397 (16.7%) eligible patients. Use of preoperative MRI more than doubled the rate of TM (odds ratio [OR] = 2.44; 95% CI, 1.58-3.77; p < 0.0001). Increasing tumor size, high nuclear grade, and age were also associated with increased rates of initial TM. Differences by age and ethnicity were observed, and significant variation in the frequency of TM was seen at the individual surgeon level (p < 0.001). Our results were similar when restricted to tumors <20 mm. CONCLUSIONS: We identified factors associated with initial TM, including preoperative MRI and individual surgeon, that contribute to the current debate about variation in use of TM for the management of breast cancer. Additional evaluation of patient understanding of surgical options and outcomes in breast cancer and the impact of the surgeon provider is warranted.

Killer cell immunoglobulin-like receptor (KIR) gene content variation in the HGDP-CEPH populations.

In the present study, we investigate patterns of variation in the KIR cluster in a large and well-characterized sample of worldwide human populations in the Human Genome Diversity Project-Centre d'Etude du Polymorphisme Humain (HGDP-CEPH) panel in order to better understand the patterns of diversity in the region. Comparison of KIR data with that from other genomic regions allows control for strictly demographic factors; over 500,000 additional genomic markers have been typed in this panel by other investigators and the data made publicly available. Presence/absence frequencies and haplotypic associations for the KIR region are analyzed in the 52 populations comprising the panel and in accordance with major world regions (Africa, Middle East, Central Asia, East Asia, Europe, Americas, and Oceania). These data represent the first overview of KIR population genetics in the well-documented HGDP-CEPH panel and suggest different evolutionary histories and recent selection in the KIR gene cluster.

Standard methods for the management of immunogenetic data.

In this chapter, we outline some basic principles for the consistent management of immunogenetic data. These include the preparation of a single master data file that can serve as the basis for all subsequent analyses, a focus on the quality and homogeneity of the data to be analyzed, the documentation of the coding systems used to represent the data, and the application of nomenclature standards specific for each immunogenetic system being evaluated. The data management principles discussed here are intended to provide a foundation for the data analysis methods detailed in Chaps. 13 and 14 . The relationship between the data management and analysis methods covered in these three chapters is illustrated in Fig. 3.The application of these data management principles is a first step toward consistent and reproducible data analyses. While it may take extra time and effort to apply them, we feel that it is better to take this approach than to assume that low data quality can be compensated for by large sample sizes.In addition to their relevance for analytical reproducibility, it is important to consider these data management principles from an ethical perspective. The reliability of the data collected and generated as part of a research study should be as important a component of the ethical review of a research application as the security of those data. Finally, in addition to ensuring the integrity of the data from collection to publication, the application of these data management principles will provide a means to foster research integrity and to improve the potential for collaborative data sharing.

Analytical methods for immunogenetic population data.

In this chapter, we describe analyses commonly applied to immunogenetic population data, along with software tools that are currently available to perform those analyses. Where possible, we focus on tools that have been developed specifically for the analysis of highly polymorphic immunogenetic data. These analytical methods serve both as a means to examine the appropriateness of a dataset for testing a specific hypothesis, as well as a means of testing hypotheses. Rather than treat this chapter as a protocol for analyzing any population dataset, each researcher and analyst should first consider their data, the possible analyses, and any available tools in light of the hypothesis being tested. The extent to which the data and analyses are appropriate to each other should be determined before any analyses are performed.

Variability in reexcision following breast conservation surgery.

CONTEXT: Health care reform calls for increasing physician accountability and transparency of outcomes. Partial mastectomy is the most commonly performed procedure for invasive breast cancer and often requires reexcision. Variability in reexcision might be reflective of the quality of care. OBJECTIVE: To assess hospital and surgeon-specific variation in reexcision rates following partial mastectomy. DESIGN, SETTING, AND PATIENTS: An observational study of breast surgery performed between 2003 and 2008 intended to evaluate variability in breast cancer surgical care outcomes and evaluate potential quality measures of breast cancer surgery. Women with invasive breast cancer undergoing partial mastectomy from 4 institutions were studied (1 university hospital [University of Vermont] and 3 large health plans [Kaiser Permanente Colorado, Group Health, and Marshfield Clinic]). Data were obtained from electronic medical records and chart abstraction of surgical, pathology, radiology, and outpatient records, including detailed surgical margin status. Logistic regression including surgeon-level random effects was used to identify predictors of reexcision. MAIN OUTCOME MEASURE: Incidence of reexcision. RESULTS: A total of 2206 women with 2220 invasive breast cancers underwent partial mastectomy and 509 patients (22.9%; 95% CI, 21.2%-24.7%) underwent reexcision (454 patients [89.2%; 95% CI, 86.5%-91.9%] had 1 reexcision, 48 [9.4%; 95% CI, 6.9%-12.0%] had 2 reexcisions, and 7 [1.4%; 95% CI, 0.4%-2.4%] had 3 reexcisions). Among all patients undergoing initial partial mastectomy, total mastectomy was performed in 190 patients (8.5%; 95% CI, 7.2%-9.5%). Reexcision rates for margin status following initial surgery were 85.9% (95% CI, 82.0%-89.8%) for initial positive margins, 47.9% (95% CI, 42.0%-53.9%) for less than 1.0 mm margins, 20.2% (95% CI, 15.3%-25.0%) for 1.0 to 1.9 mm margins, and 6.3% (95% CI, 3.2%-9.3%) for 2.0 to 2.9 mm margins. For patients with negative margins, reexcision rates varied widely among surgeons (range, 0%-70%; P = .003) and institutions (range, 1.7%-20.9%; P < .001). Reexcision rates were not associated with surgeon procedure volume after adjusting for case mix (P = .92). CONCLUSION: Substantial surgeon and institutional variation were observed in reexcision following partial mastectomy in women with invasive breast cancer.

Sequence feature variant type (SFVT) analysis of the HLA genetic association in juvenile idiopathic arthritis.

The immune response HLA class II DRB1 gene provides the major genetic contribution to Juvenile Idiopathic Arthritis (JIA), with a hierarchy of predisposing through intermediate to protective effects. With JIA, and the many other HLA associated diseases, it is difficult to identify the combinations of biologically relevant amino acid (AA) residues directly involved in disease due to the high level of HLA polymorphism, the pattern of AA variability, including varying degrees of linkage disequilibrium (LD), and the fact that most HLA variation occurs at functionally important sites. In a subset of JIA patients with the clinical phenotype oligoarticular-persistent (OP), we have applied a recently developed novel approach to genetic association analyses with genes/proteins sub-divided into biologically relevant smaller sequence features (SFs), and their "alleles" which are called variant types (VTs). With SFVT analysis, association tests are performed on variation at biologically relevant SFs based on structural (e.g., beta-strand 1) and functional (e.g., peptide binding site) features of the protein. We have extended the SFVT analysis pipeline to additionally include pairwise comparisons of DRB1 alleles within serogroup classes, our extension of the Salamon Unique Combinations algorithm, and LD patterns of AA variability to evaluate the SFVT results; all of which contributed additional complementary information. With JIA-OP, we identified a set of single AA SFs, and SFs in which they occur, particularly pockets of the peptide binding site, that account for the major disease risk attributable to HLA DRB1. These are (in numeric order): AAs 13 (pockets 4 and 6), 37 and 57 (both pocket 9), 67 (pocket 7), 74 (pocket 4), and 86 (pocket 1), and to a lesser extent 30 (pockets 6 and 7) and 71 (pockets 4, 5, and 7).

Distribution of killer cell immunoglobulin-like receptors (KIR) and their HLA-C ligands in two Iranian populations.

Killer cell immunoglobulin-like receptors (KIR) gene frequencies vary between populations and contribute to functional variation in immune responses to viruses,autoimmunity and reproductive success. This study describes the frequency distribution of 12 variable KIR genes and their HLA-C ligands in two Iranian populations who have lived for many generations in different environments:t he Azerbaijanis at high altitude and the Jonobi people at sea level. The results are compared with those published for other human populations and a large group of English Caucasians. Differences were seen in KIR and HLA-C group frequencies, in linkage disequilibrium and inhibitory/activating KIR ratios between the groups. Similarities with geographically close populations in the frequencies of the KIR A and B haplotypes and KIR AA genotype reflected their common ancestry. The extreme variability of the KIR gene family and their HLA-C ligands is highlighted and their importance in defining differences between geographically and culturally isolated communities subject to different environmental pressures who come from the same ethnic grouping.

Methods for assessing gene content diversity of KIR with examples from a global set of populations.

A number of statistical methods are widely used to describe allelic variation at specific genetic loci and its implication on the evolutionary history of these loci. Although the methods were developed primarily to study allelic variation at loci that are virtually always present in the genome, they are often applied to data of gene content variation (i.e., presence/absence of multiple homologous genes) at the killer cell immunoglobulin-like receptor (KIR) gene cluster. In this paper, we discuss methodological issues involved in the analysis of gene content variation data in the KIR region and also its covariation with polymorphism at the human leukocyte antigen class I loci, which encode ligands for KIR. A comparison of several statistical methods and measures (gene frequency, haplotype frequency, and linkage disequilibrium estimation) using the Centre d'Etude du Polymorphisme Humain data will be provided using KIR haplotypes that have been determined by segregation analysis, noting the strengths and weaknesses of the methods when only the presence/absence data is considered. Finally, application of these methods to a set of globally distributed populations is described (see Single et al., Nat Genet 39:1114-1119, 2007) in order to illustrate the challenges faced when inferring the joint effects of natural selection and demographic history on these immune-related genes.

Balancing selection and heterogeneity across the classical human leukocyte antigen loci: a meta-analytic review of 497 population studies.

This paper presents a meta-analysis of high-resolution human leukocyte antigen (HLA) allele frequency data describing 497 population samples. Most of the datasets were compiled from studies published in eight journals from 1990 to 2007; additional datasets came from the International Histocompatibility Workshops and from the AlleleFrequencies.net database. In all, these data represent approximately 66,800 individuals from throughout the world, providing an opportunity to observe trends that may not have been evident at the time the data were originally analyzed, especially with regard to the relative importance of balancing selection among the HLA loci. Population genetic measures of allele frequency distributions were summarized across populations by locus and geographic region. A role for balancing selection maintaining much of HLA variation was confirmed. Further, the breadth of this meta-analysis allowed the ranking of the HLA loci, with DQA1 and HLA-C showing the strongest balancing selection and DPB1 being compatible with neutrality. Comparisons of the allelic spectra reported by studies since 1990 indicate that most of the HLA alleles identified since 2000 are very-low-frequency alleles. The literature-based allele-count data, as well as maps summarizing the geographic distributions for each allele, are available online.

Laparoscopic approaches to resection of suspected gastric gastrointestinal stromal tumors based on tumor location.

BACKGROUND: Surgical resection of gastric gastrointestinal stromal tumor (GIST) should be optimized to achieve a negative pathologic surgical margin while limiting the extent of stomach volume loss. Careful identification of exact gastric tumor location using preoperative computed tomography (CT) scans and gastroscopy should allow for selection of a specific operative approach. METHODS: This retrospective case series involved 12 patients (7 men and 5 women; mean age, 60.5 years) with suspected gastric GIST undergoing tumor resection at Fletcher Allen Health Care, a university medical center, from January 2005 to August 2006. The main outcome measures were pathologic resection margins, operative time, estimated blood loss (EBL), morbidity, and duration of hospital stay. RESULTS: The 12 patients were separated into three groups on the basis of tumor location as follows: type 1 (fundus/greater curvature, n = 5), type 2 (prepyloric/antrum, n = 3), and type 3 (lesser curvature/perigastroesophageal junction, n = 4). Preoperative imaging (CT scan and/or endoscopy) used to identify tumor location accurately predicted the operative approach before surgery for 11 of the12 patients. The surgical approach was selected solely by tumor location as follows: type 1 (laparoscopic partial gastrectomy [LPG]), type 2 (laparoscopic distal gastrectomy [LDG]), and type 3 (laparoscopic transgastric resection [LTG]). Nine patients had a final pathologic diagnosis of GIST. The average tumor size was 4.6 cm, but this did not influence procedure selection. Histologic margins were microscopically negative in all patients. The LPG and LTG approaches had similar outcomes in terms of estimated blood loss (EBL; 80 vs 100 ml) and hospital stay (3.4 vs 3.3 days; p = 0.0198), but LTG had longer operative times (236 vs 180 min). The LDG procedure had longer operative times, greater EBL, and a longer hospital stay. The operative morbidity was 17%, and there was no operative mortality. CONCLUSION: The selection of an operative technique for resection of gastric submucosal tumors can be based on preoperative identification of tumor location, for better definition of both the extent of gastric resection and the technical complexity of the laparoscopic procedure.

Global diversity and evidence for coevolution of KIR and HLA.

The killer immunoglobulin-like receptor (KIR) gene cluster shows extensive genetic diversity, as do the HLA class I loci, which encode ligands for KIR molecules. We genotyped 1,642 individuals from 30 geographically distinct populations to examine population-level evidence for coevolution of these two functionally related but unlinked gene clusters. We observed strong negative correlations between the presence of activating KIR genes and their corresponding HLA ligand groups across populations, especially KIR3DS1 and its putative HLA-B Bw4-80I ligands (r = -0.66, P = 0.038). In contrast, we observed weak positive relationships between the various inhibitory KIR genes and their ligands. We observed a negative correlation between distance from East Africa and frequency of activating KIR genes and their corresponding ligands, suggesting a balance between selection on HLA and KIR loci. Most KIR-HLA genetic association studies indicate a primary influence of activating KIR-HLA genotypes in disease risk; concomitantly, activating receptor-ligand pairs in this study show the strongest signature of coevolution of these two complex genetic systems as compared with inhibitory receptor-ligand pairs.

Enhancing compliance with Medicare guidelines for surgical infection prevention: experience with a cross-disciplinary quality improvement team.

HYPOTHESIS: The Center for Medicare and Medicaid Services instituted standardized reporting of measures aimed at surgical infection prevention (SIP). The complexity and number of medical personnel involved in antibiotic administration requires multiple disciplines to successfully improve compliance. DESIGN: Survey study. SETTING: Tertiary care university hospital. PATIENTS: All patients undergoing the following operations from July 2004 through December 2005 were monitored for compliance with SIP: (1) coronary artery bypass graft, (2) other cardiac, (3) vascular, (4) hysterectomy, (5) colon resection, (6) hip arthroplasty, and (7) knee arthroplasty. INTERVENTION: A team including a surgeon, an anesthesiologist, nurses (preoperative, operating room, and floor), a pharmacist, a hospital infection control committee member, and quality improvement and operations specialists was created in July 2004. Hospital guidelines for SIP were defined, personnel roles defined and processes standardized, and communication/education for health care professionals was enhanced. MAIN OUTCOME MEASURES: Compliance with 3 SIP measures over 3 consecutive periods of 6 months each: (1) percentage of patients receiving antibiotics within 1 hour of incision, (2) percentage of patients with appropriately selected antibiotics, and (3) percentage of patients with antibiotics discontinued within 24 hours of operation end time. RESULTS: One thousand seventy-two patients were monitored. Measure 1 compliance improved from 72.25% to 83.78% (P<.001, Cochran-Armitage trend test); improvement or high performance (>90% compliance) was demonstrated in 5 of 7 services. Measure 2 compliance remained uniformly high (approximately 98%). Measure 3 compliance improved from 54.5% to 87.16% (P<.001); improvement was seen in 5 of 7 services. CONCLUSIONS: The clearly defined roles of a cross-disciplinary team and the process improvements discussed in this article can easily be implemented in other institutions. These elements were integral to our success in improving the timely delivery and discontinuation of prophylactic surgical antibiotics.

Diversity of MICA and linkage disequilibrium with HLA-B in two North American populations.

The MICA gene has a high degree of polymorphism. Allelic variation of MICA may influence binding of these ligands to the NK cell receptor NKG2D and may affect organ transplantation and/or disease pathogenesis. Knowledge of the population distribution of MICA alleles and their linkage disequilibrium (LD) with class I human leukocyte antigen (HLA) will enhance our understanding of the potential functional significance of the MICA polymorphism. In the present study, we characterized the MICA and HLA-B polymorphisms in two North American populations: European and African. The individual racial groups showed rather limited variation at the MICA locus, where the same set of three most common alleles, MICA*00201, *004, and *00801, account for 64 and 71% of the allele frequency in European-Americans and African-Americans, respectively. Other common alleles (allele frequency >5% in a population) include MICA*00901 and *010. MICA alleles showed strong linkage disequilibrium with HLA-B. Typically, a common MICA allele has strong LD with several HLA-B alleles, whereas most HLA-B alleles and their related serological groups are associated with a single MICA allele. The lack of evidence for an active diversification of the MICA gene after racial separation indicates an evolutionary history distinct from that of the classical HLA genes.

Signatures of demographic history and natural selection in the human major histocompatibility complex Loci.

Many lines of evidence show that several HLA loci have experienced balancing selection. However, distinguishing among demographic and selective explanations for patterns of variation observed with HLA genes remains a challenge. In this study we address this issue using data from a diverse set of human populations at six classical HLA loci and, employing a comparative genomics approach, contrast results for HLA loci to those for non-HLA markers. Using a variety of analytic methods, we confirm and extend evidence for selection acting on several HLA loci. We find that allele frequency distributions for four of the six HLA loci deviate from neutral expectations and show that this is unlikely to be explained solely by demographic factors. Other features of HLA variation are explained in part by demographic history, including decreased heterozygosity and increased LD for populations at greater distances from Africa and a similar apportionment of genetic variation for HLA loci compared to putatively neutral non-HLA loci. On the basis of contrasts among different HLA loci and between HLA and non-HLA loci, we conclude that HLA loci bear detectable signatures of both natural selection and demographic history.