Exome Sequencing of a Multigenerational Human Pedigree

Over the next few years, the efficient use of next-generation sequencing (NGS) in human genetics research will depend heavily upon the effective mechanisms for the selective enrichment of genomic regions of interest. Recently, comprehensive exome capture arrays have become available for targeting approximately 33 Mb or ∼180,000 coding exons across the human genome. Selective genomic enrichment of the human exome offers an attractive option for new experimental designs aiming to quickly identify potential disease-associated genetic variants, especially in family-based studies. We have evaluated a 2.1 M feature human exome capture array on eight individuals from a three-generation family pedigree. We were able to cover up to 98% of the targeted bases at a long-read sequence read depth of ≥3, 86% at a read depth of ≥10, and over 50% of all targets were covered with ≥20 reads. We identified up to 14,284 SNPs and small indels per individual exome, with up to 1,679 of these representing putative novel polymorphisms. Applying the conservative genotype calling approach HCDiff, the average rate of detection of a variant allele based on Illumina 1 M BeadChips genotypes was 95.2% at ≥10x sequence. Further, we propose an advantageous genotype calling strategy for low covered targets that empirically determines cut-off thresholds at a given coverage depth based on existing genotype data. Application of this method was able to detect >99% of SNPs covered ≥8x. Our results offer guidance for “real-world” applications in human genetics and provide further evidence that microarray-based exome capture is an efficient and reliable method to enrich for chromosomal regions of interest in next-generation sequencing experiments.     

Comparison of the Illumina Genome Analyzer and Roche 454 GS FLX for Resequencing of Hypertrophic Cardiomyopathy-Associated Genes

Next-generation sequencing (NGS) is widely used in biomedical research, but its adoption has been limited in molecular diagnostics. One application of NGS is the targeted resequencing of genes whose mutations lead to an overlapping clinical phenotype. This study evaluated the comparative performance of the Illumina Genome Analyzer and Roche 454 GS FLX for the resequencing of 16 genes associated with hypertrophic cardiomyopathy (HCM). Using a single human genomic DNA sample enriched by long-range PCR (LR-PCR), 40 GS FLX and 31 Genome Analyzer exon variants were identified using ≥30-fold read-coverage and ≥20% read-percentage selection criteria. Twenty-seven platform concordant variants were Sanger-confirmed. The discordant variants segregated into two categories: variants with read coverages ≥30 on one platform but <30-fold on the alternate platform and variants with read percentages ≥20% on one platform but <20% on the alternate platform. All variants with <30-fold coverage were Sanger-confirmed, suggesting that the coverage criterion of ≥30-fold is too stringent for variant discovery. The variants with <20% read percentage were identified as reference sequence based on Sanger sequencing. These variants were found in homopolymer tracts and short-read misalignments, specifically in genes with high identity. The results of the current study demonstrate the feasibility of combining LR-PCR with the Genome Analyzer or GS FLX for targeted resequencing of HCM-associated genes.     

Power to detect risk alleles using genome-wide tag SNP panels

Advances in high-throughput genotyping and the International HapMap Project have enabled association studies at the whole-genome level. We have constructed whole-genome genotyping panels of over 550,000 (HumanHap550) and 650,000 (HumanHap650Y) SNP loci by choosing tag SNPs from all populations genotyped by the International HapMap Project. These panels also contain additional SNP content in regions that have historically been overrepresented in diseases, such as nonsynonymous sites, the MHC region, copy number variant regions and mitochondrial DNA. We estimate that the tag SNP loci in these panels cover the majority of all common variation in the genome as measured by coverage of both all common HapMap SNPs and an independent set of SNPs derived from complete resequencing of genes obtained from SeattleSNPs. We also estimate that, given a sample size of 1,000 cases and 1,000 controls, these panels have the power to detect single disease loci of moderate risk (λ ~ 1.8–2.0). Relative risks as low as λ ~ 1.1–1.3 can be detected using 10,000 cases and 10,000 controls depending on the sample population and disease model. If multiple loci are involved, the power increases significantly to detect at least one locus such that relative risks 20%–35% lower can be detected with 80% power if between two and four independent loci are involved. Although our SNP selection was based on HapMap data, which is a subset of all common SNPs, these panels effectively capture the majority of all common variation and provide high power to detect risk alleles that are not represented in the HapMap data.     

Genomic DNA Enrichment Using Sequence Capture Microarrays: a Novel Approach to Discover Sequence Nucleotide Polymorphisms (SNP) in Brassica napus L

Targeted genomic selection methodologies, or sequence capture, allow for DNA enrichment and large-scale resequencing and characterization of natural genetic variation in species with complex genomes, such as rapeseed canola (Brassica napus L., AACC, 2n=38). The main goal of this project was to combine sequence capture with next generation sequencing (NGS) to discover single nucleotide polymorphisms (SNPs) in specific areas of the B. napus genome historically associated (via quantitative trait loci –QTL– analysis) to traits of agronomical and nutritional importance. A 2.1 million feature sequence capture platform was designed to interrogate DNA sequence variation across 47 specific genomic regions, representing 51.2 Mb of the Brassica A and C genomes, in ten diverse rapeseed genotypes. All ten genotypes were sequenced using the 454 Life Sciences chemistry and to assess the effect of increased sequence depth, two genotypes were also sequenced using Illumina HiSeq chemistry. As a result, 589,367 potentially useful SNPs were identified. Analysis of sequence coverage indicated a four-fold increased representation of target regions, with 57% of the filtered SNPs falling within these regions. Sixty percent of discovered SNPs corresponded to transitions while 40% were transversions. Interestingly, fifty eight percent of the SNPs were found in genic regions while 42% were found in intergenic regions. Further, a high percentage of genic SNPs was found in exons (65% and 64% for the A and C genomes, respectively). Two different genotyping assays were used to validate the discovered SNPs. Validation rates ranged from 61.5% to 84% of tested SNPs, underpinning the effectiveness of this SNP discovery approach. Most importantly, the discovered SNPs were associated with agronomically important regions of the B. napus genome generating a novel data resource for research and breeding this crop species.     

NotI flanking sequences: a tool for gene discovery and verification of the human genome

A set of 22 551 unique human NotI flanking sequences (16.2 Mb) was generated. More than 40% of the set had regions with significant similarity to known proteins and expressed sequences. The data demonstrate that regions flanking NotI sites are less likely to form nucleosomes efficiently and resemble promoter regions. The draft human genome sequence contained 55.7% of the NotI flanking sequences, Celera’s database contained matches to 57.2% of the clones and all public databases (including non-human and previously sequenced NotI flanks) matched 89.2% of the NotI flanking sequences (identity ≥90% over at least 50 bp, data from December 2001). The data suggest that the shotgun sequencing approach used to generate the draft human genome sequence resulted in a bias against cloning and sequencing of NotI flanks. A rough estimation (based primarily on chromosomes 21 and 22) is that the human genome contains 15 000–20 000 NotI sites, of which 6000–9000 are unmethylated in any particular cell. The results of the study suggest that the existing tools for computational determination of CpG islands fail to identify a significant fraction of functional CpG islands, and unmethylated DNA stretches with a high frequency of CpG dinucleotides can be found even in regions with low CG content.     

Comparison of Commercially Available Target Enrichment Methods for Next-Generation Sequencing

Isolating high-priority segments of genomes greatly enhances the efficiency of next-generation sequencing (NGS) by allowing researchers to focus on their regions of interest. For the 2010–11 DNA Sequencing Research Group (DSRG) study, we compared outcomes from two leading companies, Agilent Technologies (Santa Clara, CA, USA) and Roche NimbleGen (Madison, WI, USA), which offer custom-targeted genomic enrichment methods. Both companies were provided with the same genomic sample and challenged to capture identical genomic locations for DNA NGS. The target region totaled 3.5 Mb and included 31 individual genes and a 2-Mb contiguous interval. Each company was asked to design its best assay, perform the capture in replicates, and return the captured material to the DSRG-participating laboratories. Sequencing was performed in two different laboratories on Genome Analyzer IIx systems (Illumina, San Diego, CA, USA). Sequencing data were analyzed for sensitivity, specificity, and coverage of the desired regions. The success of the enrichment was highly dependent on the design of the capture probes. Overall, coverage variability was higher for the Agilent samples. As variant discovery is the ultimate goal for a typical targeted sequencing project, we compared samples for their ability to sequence single-nucleotide polymorphisms (SNPs) as a test of the ability to capture both chromosomes from the sample. In the targeted regions, we detected 2546 SNPs with the NimbleGen samples and 2071 with Agilent''s. When limited to the regions that both companies included as baits, the number of SNPs was ∼1000 for each, with Agilent and NimbleGen finding a small number of unique SNPs not found by the other.     

An evaluation of the performance of tag SNPs derived from HapMap in a Caucasian population

The Haplotype Map (HapMap) project recently generated genotype data for more than 1 million single-nucleotide polymorphisms (SNPs) in four population samples. The main application of the data is in the selection of tag single-nucleotide polymorphisms (tSNPs) to use in association studies. The usefulness of this selection process needs to be verified in populations outside those used for the HapMap project. In addition, it is not known how well the data represent the general population, as only 90–120 chromosomes were used for each population and since the genotyped SNPs were selected so as to have high frequencies. In this study, we analyzed more than 1,000 individuals from Estonia. The population of this northern European country has been influenced by many different waves of migrations from Europe and Russia. We genotyped 1,536 randomly selected SNPs from two 500-kbp ENCODE regions on Chromosome 2. We observed that the tSNPs selected from the CEPH (Centre d'Etude du Polymorphisme Humain) from Utah (CEU) HapMap samples (derived from US residents with northern and western European ancestry) captured most of the variation in the Estonia sample. (Between 90% and 95% of the SNPs with a minor allele frequency of more than 5% have an r² of at least 0.8 with one of the CEU tSNPs.) Using the reverse approach, tags selected from the Estonia sample could almost equally well describe the CEU sample. Finally, we observed that the sample size, the allelic frequency, and the SNP density in the dataset used to select the tags each have important effects on the tagging performance. Overall, our study supports the use of HapMap data in other Caucasian populations, but the SNP density and the bias towards high-frequency SNPs have to be taken into account when designing association studies.     

Copy number variants and common disorders: filling the gaps and exploring complexity in genome-wide association studies

Genome-wide association scans (GWASs) using single nucleotide polymorphisms (SNPs) have been completed successfully for several common disorders and have detected over 30 new associations. Considering the large sample sizes and genome-wide SNP coverage of the scans, one might have expected many of the common variants underpinning the genetic component of various disorders to have been identified by now. However, these studies have not evaluated the contribution of other forms of genetic variation, such as structural variation, mainly in the form of copy number variants (CNVs). Known CNVs account for over 15% of the assembled human genome sequence. Since CNVs are not easily tagged by SNPs, might have a wide range of copy number variability, and often fall in genomic regions not well covered by whole-genome arrays or not genotyped by the HapMap project, current GWASs have largely missed the contribution of CNVs to complex disorders. In fact, some CNVs have already been reported to show association with several complex disorders using candidate gene/region approaches, underpinning the importance of regions not investigated in current GWASs. This reveals the need for new generation arrays (some already in the market) and the use of tailored approaches to explore the full dimension of genome variability beyond the single nucleotide scale.     

Comparison and quantitative verification of mapping algorithms for whole-genome bisulfite sequencing

Coupling bisulfite conversion with next-generation sequencing (Bisulfite-seq) enables genome-wide measurement of DNA methylation, but poses unique challenges for mapping. However, despite a proliferation of Bisulfite-seq mapping tools, no systematic comparison of their genomic coverage and quantitative accuracy has been reported. We sequenced bisulfite-converted DNA from two tissues from each of two healthy human adults and systematically compared five widely used Bisulfite-seq mapping algorithms: Bismark, BSMAP, Pash, BatMeth and BS Seeker. We evaluated their computational speed and genomic coverage and verified their percentage methylation estimates. With the exception of BatMeth, all mappers covered >70% of CpG sites genome-wide and yielded highly concordant estimates of percentage methylation (r² ≥ 0.95). Fourfold variation in mapping time was found between BSMAP (fastest) and Pash (slowest). In each library, 8–12% of genomic regions covered by Bismark and Pash were not covered by BSMAP. An experiment using simulated reads confirmed that Pash has an exceptional ability to uniquely map reads in genomic regions of structural variation. Independent verification by bisulfite pyrosequencing generally confirmed the percentage methylation estimates by the mappers. Of these algorithms, Bismark provides an attractive combination of processing speed, genomic coverage and quantitative accuracy, whereas Pash offers considerably higher genomic coverage.     

Single nucleotide polymorphisms (SNPs) that map to gaps in the human SNP map

An international effort is underway to generate a comprehensive haplotype map (HapMap) of the human genome represented by an estimated 300000 to 1 million ‘tag’ single nucleotide polymorphisms (SNPs). Our analysis indicates that the current human SNP map is not sufficiently dense to support the HapMap project. For example, 24.6% of the genome currently lacks SNPs at the minimal density and spacing that would be required to construct even a conservative tag SNP map containing 300 000 SNPs. In an effort to improve the human SNP map, we identified 140 696 additional SNP candidates using a new bioinformatics pipeline. Over 51 000 of these SNPs mapped to the largest gaps in the human SNP map, leading to significant improvements in these regions. Our SNPs will be immediately useful for the HapMap project, and will allow for the inclusion of many additional genomic intervals in the final HapMap. Nevertheless, our results also indicate that additional SNP discovery projects will be required both to define the haplotype architecture of the human genome and to construct comprehensive tag SNP maps that will be useful for genetic linkage studies in humans.     

Draft Sequences of the Radish (Raphanus sativus L.) Genome

Radish (Raphanus sativus L., n = 9) is one of the major vegetables in Asia. Since the genomes of Brassica and related species including radish underwent genome rearrangement, it is quite difficult to perform functional analysis based on the reported genomic sequence of Brassica rapa. Therefore, we performed genome sequencing of radish. Short reads of genomic sequences of 191.1 Gb were obtained by next-generation sequencing (NGS) for a radish inbred line, and 76,592 scaffolds of ≥300 bp were constructed along with the bacterial artificial chromosome-end sequences. Finally, the whole draft genomic sequence of 402 Mb spanning 75.9% of the estimated genomic size and containing 61,572 predicted genes was obtained. Subsequently, 221 single nucleotide polymorphism markers and 768 PCR-RFLP markers were used together with the 746 markers produced in our previous study for the construction of a linkage map. The map was combined further with another radish linkage map constructed mainly with expressed sequence tag-simple sequence repeat markers into a high-density integrated map of 1,166 cM with 2,553 DNA markers. A total of 1,345 scaffolds were assigned to the linkage map, spanning 116.0 Mb. Bulked PCR products amplified by 2,880 primer pairs were sequenced by NGS, and SNPs in eight inbred lines were identified.     

Targeted high throughput sequencing of a cancer-related exome subset by specific sequence capture with a fully automated microarray platform

Sequence capture methods for targeted next generation sequencing promise to massively reduce cost of genomics projects compared to untargeted sequencing. However, evaluated capture methods specifically dedicated to biologically relevant genomic regions are rare. Whole exome capture has been shown to be a powerful tool to discover the genetic origin of disease and provides a reduction in target size and thus calculative sequencing capacity of > 90-fold compared to untargeted whole genome sequencing. For further cost reduction, a valuable complementing approach is the analysis of smaller, relevant gene subsets but involving large cohorts of samples. However, effective adjustment of target sizes and sample numbers is hampered by the limited scalability of enrichment systems. We report a highly scalable and automated method to capture a 480 Kb exome subset of 115 cancer-related genes using microfluidic DNA arrays. The arrays are adaptable from 125 Kb to 1 Mb target size and/or one to eight samples without barcoding strategies, representing a further 26 – 270-fold reduction of calculative sequencing capacity compared to whole exome sequencing. Illumina GAII analysis of a HapMap genome enriched for this exome subset revealed a completeness of > 96%. Uniformity was such that > 68% of exons had at least half the median depth of coverage. An analysis of reference SNPs revealed a sensitivity of up to 93% and a specificity of 98.2% or higher.     

Whole genome sequencing of the black grouse (Tetrao tetrix): reference guided assembly suggests faster-Z and MHC evolution

Background

The different regions of a genome do not evolve at the same rate. For example, comparative genomic studies have suggested that the sex chromosomes and the regions harbouring the immune defence genes in the Major Histocompatability Complex (MHC) may evolve faster than other genomic regions. The advent of the next generation sequencing technologies has made it possible to study which genomic regions are evolutionary liable to change and which are static, as well as enabling an increasing number of genome studies of non-model species. However, de novo sequencing of the whole genome of an organism remains non-trivial. In this study, we present the draft genome of the black grouse, which was developed using a reference-guided assembly strategy.

Results

We generated 133 Gbp of sequence data from one black grouse individual by the SOLiD platform and used a combination of de novo assembly and chicken reference genome mapping to assemble the reads into 4572 scaffolds with a total length of 1022 Mb. The draft genome well covers the main chicken chromosomes 1 ~ 28 and Z which have a total length of 1001 Mb. The draft genome is fragmented, but has a good coverage of the homologous chicken genes. Especially, 33.0% of the coding regions of the homologous genes have more than 90% proportion of their sequences covered. In addition, we identified ~1 M SNPs from the genome and identified 106 genomic regions which had a high nucleotide divergence between black grouse and chicken or between black grouse and turkey.

Conclusions

Our results support the hypothesis that the chromosome X (Z) evolves faster than the autosomes and our data are consistent with the MHC regions being more liable to change than the genome average. Our study demonstrates how a moderate sequencing effort can be combined with existing genome references to generate a draft genome for a non-model species.

Electronic supplementary material

The online version of this article (doi:10.1186/1471-2164-15-180) contains supplementary material, which is available to authorized users.     

Metabolic potential of a single cell belonging to one of the most abundant lineages in freshwater bacterioplankton

Actinobacteria within the acI lineage are often numerically dominating in freshwater ecosystems, where they can account for >50% of total bacteria in the surface water. However, they remain uncultured to date. We thus set out to use single-cell genomics to gain insights into their genetic make-up, with the aim of learning about their physiology and ecological niche. A representative from the highly abundant acI-B1 group was selected for shotgun genomic sequencing. We obtained a draft genomic sequence in 75 larger contigs (sum=1.16 Mb), with an unusually low genomic G+C mol% (∼42%). Actinobacteria core gene analysis suggests an almost complete genome recovery. We found that the acI-B1 cell had a small genome, with a rather low percentage of genes having no predicted functions (∼15%) as compared with other cultured and genome-sequenced microbial species. Our metabolic reconstruction hints at a facultative aerobe microorganism with many transporters and enzymes for pentoses utilization (for example, xylose). We also found an actinorhodopsin gene that may contribute to energy conservation under unfavorable conditions. This project reveals the metabolic potential of a member of the global abundant freshwater Actinobacteria.     

Comparison and evaluation of two exome capture kits and sequencing platforms for variant calling

Background

To promote the clinical application of next-generation sequencing, it is important to obtain accurate and consistent variants of target genomic regions at low cost. Ion Proton, the latest updated semiconductor-based sequencing instrument from Life Technologies, is designed to provide investigators with an inexpensive platform for human whole exome sequencing that achieves a rapid turnaround time. However, few studies have comprehensively compared and evaluated the accuracy of variant calling between Ion Proton and Illumina sequencing platforms such as HiSeq 2000, which is the most popular sequencing platform for the human genome. The Ion Proton sequencer combined with the Ion TargetSeq™ Exome Enrichment Kit together make up TargetSeq-Proton, whereas SureSelect-Hiseq is based on the Agilent SureSelect Human All Exon v4 Kit and the HiSeq 2000 sequencer.

Results

Here, we sequenced exonic DNA from four human blood samples using both TargetSeq-Proton and SureSelect-HiSeq. We then called variants in the exonic regions that overlapped between the two exome capture kits (33.6 Mb). The rates of shared variant loci called by two sequencing platforms were from 68.0 to 75.3 % in four samples, whereas the concordance of co-detected variant loci reached 99 %. Sanger sequencing validation revealed that the validated rate of concordant single nucleotide polymorphisms (SNPs) (91.5 %) was higher than the SNPs specific to TargetSeq-Proton (60.0 %) or specific to SureSelect-HiSeq (88.3 %). With regard to 1-bp small insertions and deletions (InDels), the Sanger sequencing validated rates of concordant variants (100.0 %) and SureSelect-HiSeq-specific (89.6 %) were higher than those of TargetSeq-Proton-specific (15.8 %).

Conclusions

In the sequencing of exonic regions, a combination of using of two sequencing strategies (SureSelect-HiSeq and TargetSeq-Proton) increased the variant calling specificity for concordant variant loci and the sensitivity for variant loci called by any one platform. However, for the sequencing of platform-specific variants, the accuracy of variant calling by HiSeq 2000 was higher than that of Ion Proton, specifically for the InDel detection. Moreover, the variant calling software also influences the detection of SNPs and, specifically, InDels in Ion Proton exome sequencing.

Electronic supplementary material

The online version of this article (doi:10.1186/s12864-015-1796-6) contains supplementary material, which is available to authorized users.     

Oceanic Metagenomics: The Sorcerer II Global Ocean Sampling Expedition: Northwest Atlantic through Eastern Tropical Pacific

The world's oceans contain a complex mixture of micro-organisms that are for the most part, uncharacterized both genetically and biochemically. We report here a metagenomic study of the marine planktonic microbiota in which surface (mostly marine) water samples were analyzed as part of the Sorcerer II Global Ocean Sampling expedition. These samples, collected across a several-thousand km transect from the North Atlantic through the Panama Canal and ending in the South Pacific yielded an extensive dataset consisting of 7.7 million sequencing reads (6.3 billion bp). Though a few major microbial clades dominate the planktonic marine niche, the dataset contains great diversity with 85% of the assembled sequence and 57% of the unassembled data being unique at a 98% sequence identity cutoff. Using the metadata associated with each sample and sequencing library, we developed new comparative genomic and assembly methods. One comparative genomic method, termed “fragment recruitment,” addressed questions of genome structure, evolution, and taxonomic or phylogenetic diversity, as well as the biochemical diversity of genes and gene families. A second method, termed “extreme assembly,” made possible the assembly and reconstruction of large segments of abundant but clearly nonclonal organisms. Within all abundant populations analyzed, we found extensive intra-ribotype diversity in several forms: (1) extensive sequence variation within orthologous regions throughout a given genome; despite coverage of individual ribotypes approaching 500-fold, most individual sequencing reads are unique; (2) numerous changes in gene content some with direct adaptive implications; and (3) hypervariable genomic islands that are too variable to assemble. The intra-ribotype diversity is organized into genetically isolated populations that have overlapping but independent distributions, implying distinct environmental preference. We present novel methods for measuring the genomic similarity between metagenomic samples and show how they may be grouped into several community types. Specific functional adaptations can be identified both within individual ribotypes and across the entire community, including proteorhodopsin spectral tuning and the presence or absence of the phosphate-binding gene PstS.     

Bisulfite genomic sequencing: systematic investigation of critical experimental parameters

Bisulfite genomic sequencing is the method of choice for the generation of methylation maps with single-base resolution. The method is based on the selective deamination of cytosine to uracil by treatment with bisulfite and the sequencing of subsequently generated PCR products. In contrast to cytosine, 5-methylcytosine does not react with bisulfite and can therefore be distinguished. In order to investigate the potential for optimization of the method and to determine the critical experimental parameters, we determined the influence of incubation time and incubation temperature on the deamination efficiency and measured the degree of DNA degradation during the bisulfite treatment. We found that maximum conversion rates of cytosine occurred at 55°C (4–18 h) and 95°C (1 h). Under these conditions at least 84–96% of the DNA is degraded. To study the impact of primer selection, homologous DNA templates were constructed possessing cytosine-containing and cytosine-free primer binding sites, respectively. The recognition rates for cytosine (≥97%) and 5-methylcytosine (≥94%) were found to be identical for both templates.     

Genome-wide identification of human functional DNA using a neutral indel model

It has become clear that a large proportion of functional DNA in the human genome does not code for protein. Identification of this non-coding functional sequence using comparative approaches is proving difficult and has previously been thought to require deep sequencing of multiple vertebrates. Here we introduce a new model and comparative method that, instead of nucleotide substitutions, uses the evolutionary imprint of insertions and deletions (indels) to infer the past consequences of selection. The model predicts the distribution of indels under neutrality, and shows an excellent fit to human–mouse ancestral repeat data. Across the genome, many unusually long ungapped regions are detected that are unaccounted for by the neutral model, and which we predict to be highly enriched in functional DNA that has been subject to purifying selection with respect to indels. We use the model to determine the proportion under indel-purifying selection to be between 2.56% and 3.25% of human euchromatin. Since annotated protein-coding genes comprise only 1.2% of euchromatin, these results lend further weight to the proposition that more than half the functional complement of the human genome is non-protein-coding. The method is surprisingly powerful at identifying selected sequence using only two or three mammalian genomes. Applying the method to the human, mouse, and dog genomes, we identify 90 Mb of human sequence under indel-purifying selection, at a predicted 10% false-discovery rate and 75% sensitivity. As expected, most of the identified sequence represents unannotated material, while the recovered proportions of known protein-coding and microRNA genes closely match the predicted sensitivity of the method. The method's high sensitivity to functional sequence such as microRNAs suggest that as yet unannotated microRNA genes are enriched among the sequences identified. Futhermore, its independence of substitutions allowed us to identify sequence that has been subject to heterogeneous selection, that is, sequence subject to both positive selection with respect to substitutions and purifying selection with respect to indels. The ability to identify elements under heterogeneous selection enables, for the first time, the genome-wide investigation of positive selection on functional elements other than protein-coding genes.