Tagged with #haplotypecaller
7 documentation articles | 4 announcements | 0 forum discussions

Comments (0)

This document describes the methods involved in variant calling as performed by the HaplotypeCaller. Please note that this document is currently incomplete and will be updated and/or corrected over the course of the next few days.


The core operations performed by HaplotypeCaller can be grouped into these major steps:

1. Define active regions. The program determines which regions of the genome it needs to operate on, based on the presence of significant evidence for variation.

2. Determine haplotypes by re-assembly of the active region. For each ActiveRegion, the program builds a De Bruijn-like graph to reassemble the ActiveRegion and identifies what are the possible haplotypes present in the data. The program then realigns each haplotype against the reference haplotype using the Smith-Waterman algorithm in order to identify potentially variant sites.

3. Determine likelihoods of the haplotypes given the read data. For each ActiveRegion, the program performs a pairwise alignment of each read against each haplotype using the PairHMM algorithm. This produces a matrix of likelihoods of haplotypes given the read data. These likelihoods are then marginalized to obtain the likelihoods of alleles per read for each potentially variant site.

4. Assign sample genotypes. For each potentially variant site, the program applies Bayes’ rule, using the likelihoods of alleles given the read data to calculate the posterior likelihoods of each genotype per sample given the read data observed for that sample. The most likely genotype is then assigned to the sample.

1. Define active regions

In this first step, the program traverses the sequencing data to identify regions of the genomes in which the samples being analyzed show substantial evidence of variation relative to the reference. The resulting areas are defined as “active regions”, and will be passed on to the next step. Areas that do not show any variation beyond the expected levels of background noise will be skipped in the next step. This aims to accelerate the analysis by not wasting time performing reassembly on regions that are identical to the reference anyway.

To define these active regions, the program operates in three phases. First, it computes an activity score for each individual genome position, yielding the raw activity profile, which is a wave function of activity per position. Then, it applies a smoothing algorithm to the raw profile, which is essentially a sort of averaging process, to yield the actual activity profile. Finally, it identifies local maxima where the activity profile curve rises above the preset activity threshold, and defines appropriate intervals to encompass the active profile within the preset size constraints. For more details on how the activity profile is computed and processed, as well as what options are available to modify the active region parameters, please see this method article.

Note that the process for determining active region intervals is modified slightly when HaplotypeCaller is run in one of the special modes, e.g. the reference confidence mode (-ERC GVCF or ERC BP_RESOLUTION), Genotype Given Alleles (-gt_mode GENOTYPE_GIVEN_ALLELES) or when active regions are triggered using advanced arguments such as -allelesTrigger, --forceActive or --activeRegionIn. This is covered in the method article referenced above.

Once this process is complete, the program applies a few post-processing steps to finalize the the active regions (see detailed doc above). The final output of this process is a list of intervals corresponding to the active regions which will be processed in the next step.

2. Determine haplotypes by re-assembly of the active region.

The goal of this step is to reconstruct the possible sequences of the real physical segments of DNA present in the original sample organism. To do this, the program goes through each active region and uses the input reads that mapped to that region to construct complete sequences covering its entire length, which are called haplotypes. This process will typically generate several different possible haplotypes for each active region due to:

  • real diversity on polyploid (including CNV) or multi-sample data
  • possible allele combinations between variant sites that are not totally linked within the active region
  • sequencing and mapping errors

In order to generate a list of possible haplotypes, the program first builds an assembly graph for the active region using the reference sequence as a template. Then, it takes each read in turn and attempts to match it to a segment of the graph. Whenever portions of a read do not match the local graph, the program adds new nodes to the graph to account for the mismatches. After this process has been repeated with many reads, it typically yields a complex graph with many possible paths. However, because the program keeps track of how many reads support each path segment, we can select only the most likely (well-supported) paths. These likely paths are then used to build the haplotype sequences which will be used for scoring and genotyping in the next step.

The assembly and haplotype determination procedure is described in full detail in this method article.

Once the haplotypes have been determined, each one is realigned against the original reference sequence in order to identify potentially variant sites. This produces the set of sites that will be processed in the next step. A subset of these sites will eventually be emitted as variant calls to the output VCF.

3. Determine likelihoods of the haplotypes given the read data

[in progress]

4. Assign sample genotypes

[in progress]

Comments (0)

This document describes the procedure used by HaplotypeCaller to define ActiveRegions on which to operate as a prelude to variant calling. For more context information on how this fits into the overall HaplotypeCaller method, please see the more general HaplotypeCaller documentation.


To define active regions, the HaplotypeCaller operates in three phases. First, it computes an activity score for each individual genome position, yielding the raw activity profile, which is a wave function of activity per position. Then, it applies a smoothing algorithm to the raw profile, which is essentially a sort of averaging process, to yield the actual activity profile. Finally, it identifies local maxima where the activity profile curve rises above the preset activity threshold, and defines appropriate intervals to encompass the active profile within the preset size constraints.

1. Calculating the raw activity profile

Active regions are determined by calculating a profile function that characterizes “interesting” regions likely to contain variants. The raw profile is first calculated locus by locus.

In the normal case (no special mode is enabled) the per-position score is the probability that the position contains a variant as calculated using the reference-confidence model applied to the original alignment.

If using the mode for genotyping given alleles (GGA) or the advanced-level flag -useAlleleTrigger, and the site is overlapped by an allele in the VCF file provided through the -alleles argument, the score is set to 1. If the position is not covered by a provided allele, the score is set to 0.

This operation give us a single raw value for each position on the genome (or within the analysis intervals requested using the -L argument).

2. Smoothing the activity profile

The final profile is calculated by smoothing this initial raw profile following three steps. The first two steps consist in spreading individual position raw profile values to contiguous bases. As a result each position will have more than one raw profile value that are added up in the third and last step to obtain a final unique and smoothed value per position.

  1. Unless one of the special modes is enabled (GGA or allele triggering), the position profile value will be copied over to adjacent regions if enough high quality soft-clipped bases immediately precede or follow that position in the original alignment. At time of writing, high-quality soft-clipped bases are those with quality score of Q29 or more. We consider that there are enough of such a soft-clips when the average number of high quality bases per soft-clip is 7 or more. In this case the site profile value is copied to all bases within a radius of that position as large as the average soft-clip length without exceeding a maximum of 50bp.

  2. Each profile value is then divided and spread out using a Gaussian kernel covering up to 50bp radius centered at its current position with a standard deviation, or sigma, set using the -bandPassSigma argument (current default is 17 bp). The larger the sigma, the broader the spread will be.

  3. For each position, the final smoothed value is calculated as the sum of all its profile values after steps 1 and 2.

3. Setting the ActiveRegion thresholds and intervals

The resulting profile line is cut in regions where it crosses the non-active to active threshold (currently set to 0.002). Then we make some adjustments to these boundaries so that those regions that are to be considered active, with a profile running over that threshold, fall within the minimum (fixed to 50bp) and maximum region size (customizable using -activeRegionMaxSize).

  • If the region size falls within the limits we leave it untouched (it's good to go).

  • If the region size is shorter than the minimum, it is greedily extended forward ignoring that cut point and we come back to step 1. Only if this is not possible because we hit a hard-limit (end of the chromosome or requested analysis interval) we will accept the small region as it is.

  • If it is too long, we find the lowest local minimum between the maximum and minimum region size. A local minimum is a profile value preceded by a large one right up-stream (-1bp) and an equal or larger value down-stream (+1bp). In case of a tie, the one further downstream takes precedence. If there is no local minimum we simply force the cut so that the region has the maximum active region size.

Of the resulting regions, those with a profile that runs over this threshold are considered active regions and progress to variant discovery and or calling whereas regions whose profile runs under the threshold are considered inactive regions and are discarded except if we are running HC in reference confidence mode.

There is a final post-processing step to clean up and trim the ActiveRegion:

  • Remove bases at each end of the read (hard-clipping) until there a base with a call quality equal or greater than minimum base quality score (customizable parameter -mbq, 10 by default).

  • Include or exclude remaining soft-clipped ends. Soft clipped ends will be used for assembly and calling unless the user has requested their exclusion (using -dontUseSoftClippedBases), if the read and its mate map to the same chromosome, and if they are in the correct standard orientation (i.e. LR and RL).

  • Clip off adaptor sequences of the read if present.

  • Discard all reads that no longer overlap with the ActiveRegion after the trimming operations described above.

  • Downsample remaining reads to a maximum of 1000 reads per sample, but respecting a minimum of 5 reads starting per position. This is performed after any downsampling by the traversal itself (-dt, -dfrac, -dcov etc.) and cannot be overriden from the command line.

Comments (0)

This document details the procedure used by HaplotypeCaller to re-assemble read data and determine candidate haplotypes as a prelude to variant calling. For more context information on how this fits into the overall HaplotypeCaller method, please see the more general HaplotypeCaller documentation.

Note that we are still working on producing figures to complement the text. We will update this document as soon as the figures are ready. Note also that this is a provisional document and some final corrections may be made for accuracy and/or completeness. Feedback is most welcome!


The previous step produced a list of ActiveRegions that showed some evidence of possible variation (see step 1 documentation). Now, we need to process each Active Region in order to generate a list of possible haplotypes based on the sequence data we have for that region.

To do so, the program first builds an assembly graph for each active region (determined in the previous step) using the reference sequence as a template. Then, it takes each read in turn and attempts to match it to a segment of the graph. Whenever portions of a read do not match the local graph, the program adds new nodes to the graph to account for the mismatches. After this process has been repeated with many reads, it typically yields a complex graph with many possible paths. However, because the program keeps track of how many reads support each path segment, we can select only the most likely (well-supported) paths. These likely paths are then used to build the haplotype sequences which will be used to call variants and assign per-sample genotypes in the next steps.

1. Reference graph assembly

First, we construct the reference assembly graph, which starts out as a simple directed DeBruijn graph. This involves decomposing the reference sequence into a succession of kmers (pronounced "kay-mers"), which are small sequence subunits that are k bases long. Each kmer sequence overlaps the previous kmer by k-1 bases. The resulting graph can be represented as a series of nodes and connecting edges indicating the sequential relationship between the adjacent bases. At this point, all the connecting edges have a weight of 0.

In addition to the graph, we also build a hash table of unique kmers, which we use to keep track of the position of nodes in the graph. At the beginning, the hash table only contains unique kmers found in the reference sequence, but we will add to it in the next step.

A note about kmer size: by default, the program will attempt to build two separate graphs, using kmers of 10 and 25 bases in size, respectively, but other kmer sizes can be specified from the command line with the -kmerSize argument. The final set of haplotypes will be selected from the union of the graphs obtained using each k.

2. Threading reads through the graph

This is where our simple reference graph turns into a read-threading graph, so-called because we're going to take each read in turn and try to match it to a path in the graph.

We start with the first read and compare its first kmer to the hash table to find if it has a match. If there is a match, we look up its position in the reference graph and record that position. If there is no match, we consider that it is a new unique kmer, so we add that unique kmer to the hash table and add a new node to the graph. In both cases, we then move on and repeat the process with the next kmer in the read until we reach the end of the read.

When two consecutive kmers in a read belong to two nodes that were already connected by an edge in the graph, we increase the weight of that edge by 1. If the two nodes were not connected yet, we add a new edge to the graph with a starting weight of 1. As we repeat the process on each read in turn, edge weights will accumulate along the paths that are best supported by the read data, which will help us select the most likely paths later on.

Note on graph complexity, cycles and non-unique kmers

For this process to work properly, we need the graph to be sufficiently complex (where the number of non-unique k-mers is less that 4-fold the number of unique kmers found in the data) and without cycles. In certain genomic regions where there are a lot of repeated sequences, these conditions may not be met, because repeats cause cycles and diminish the number of available unique kmers. If none of the kmer sizes provided results in a viable graph (complex enough and without cycles) the program will automatically try the operation again with larger kmer sizes. Specifically, we take the largest k provided by the user (or by the default settings) and increase it by 10 bases. If no viable graph can be obtained after iterating over increased kmer sizes 6 times, we give up and skip the active region entirely.

3. Graph refinement

Once all the reads have been threaded through the graph, we need to clean it up a little. The main cleaning-up operation is called pruning (like the gardening technique). The goal of the pruning operation is to remove noise due to errors. The basic idea is that sections of the graph that are supported by very few reads are most probably the result of stochastic errors, so we are going to remove any sections that are supported by fewer than a certain threshold number of reads. By default the threshold value is 2, but this can be controlled from the command line using the -minPruning argument. In practice, this means that linear chains in the graph (linear sequence of vertices and edges without any branching) where all edges have fewer than 2 supporting reads will be removed. Increasing the threshold value will lead to faster processing and higher specificity, but will decrease sensitivity. Decreasing this value will do the opposite, decreasing specificity but increasing sensitivity.

Note that if you are calling multiple samples together, the program also looks at how many of the samples support each segment, and only prunes segments for which fewer than a certain number of samples have the minimum required number of supporting reads. By default this sample number is 1, so as long as one sample in the cohort passes the pruning threshold, the segment will NOT be pruned. This is designed to avoid losing singletons (variants that are unique to a single sample in a cohort). This parameter can also be controlled from the command line using the -minPruningSamples argument, but keep in mind that increasing the default value may lead to decreased sensitivity.

At this stage, the program can also perform graph refinement operations that are specific to certain special modes. For example, when HaplotypeCaller is run on RNAseq, it will recover dangling heads and tails from the splice junctions to compensate for issues that are related to limitations in graph assembly of that data type.

4. Select best haplotypes

Now that the graph is all cleaned up, the program builds haplotype sequences by traversing all possible paths in the graph and calculates a likelihood score for each one. This score is calculated as the product of transition probabilities of the path edges, where the transition probability of an edge is computed as the number of reads supporting that edge divided by the sum of the support of all edges that share that same source vertex.

In order to limit the amount of computation needed for the next step, we limit the number of haplotypes that will be considered for each value of k (remember that the program builds graphs for multiple kmer sizes). This is easy to do since we conveniently have scores for each haplotype; all we need to do is select the N haplotypes with the best scores. By default that number is very generously set to 128 (so the program would proceed to the next step with up to 128 haplotypes per value of k) but this can be adjusted from the command line using the -maxNumHaplotypesInPopulation argument. You would mainly want to decrease this number in order to improve speed; increasing that number would rarely be reasonable, if ever.

5. Identify potential variation sites

Once we have a list of plausible haplotypes, we perform a Smith-Waterman alignment (SWA) of each haplotype to the original reference sequence across the active region in order to reconstruct a CIGAR string for the haplotype. Note that indels will be left-aligned; that is, their start position will be set as the leftmost position possible.

This finally yields the potential variation sites that will be put through the variant modeling step next, bringing us back to the "classic" variant calling methods (as used by GATK's UnifiedGenotyper and Samtools' mpileup). Note that this list of candidate sites is essentially a super-set of what will eventually be the final set of called variants. Every site that will be called variant is in the super-set, but not every site that is in the super-set will be called variant.

Comments (0)

This document describes the reference confidence model applied by HaplotypeCaller to generate genomic VCFs (gVCFS), invoked by -ERC GVCF or -ERC BP_RESOLUTION (see the FAQ on gVCFs for format details).

Please note that this document may be expanded with more detailed information in the near future.

How it works

The mode works by assembling the reads to create potential haplotypes, realigning the reads to their most likely haplotypes, and then projecting these reads back onto the reference sequence via their haplotypes to compute alignments of the reads to the reference. For each position in the genome we have either a non-reference call (via the standard calling mechanism) or we can estimate the chance that some (unknown) non-reference allele is segregating at this position by examining the realigned reads that span the reference base. At this base we perform two calculations:

  • Estimate the confidence that no SNP exists at the site by contrasting all reads with the ref base vs all reads with any non-reference base.
  • Estimate the confidence that no indel of size < X (determined by command line parameter) could exist at this site by calculating the number of reads that provide evidence against such an indel, and from this value estimate the chance that we would not have seen the allele confidently.

Based on this, we emit the genotype likelihoods (PL) and compute the GQ (from the PLs) for the least confidence of these two models.

We use a symbolic allele, <NON_REF>, to indicate that the site is homozygous reference, and because we have an ALT allele we can provide allele-specific AD and PL field values.

For details of the gVCF format, please see the document that explains what is a gVCF.

Comments (1)


GVCF stands for Genomic VCF. A GVCF is a kind of VCF, so the basic format specification is the same as for a regular VCF (see the spec documentation here), but a Genomic VCF contains extra information.

This document explains what that extra information is and how you can use it to empower your variants analyses.

Important caveat

What we're covering here is strictly limited to GVCFs produced by HaplotypeCaller in GATK versions 3.0 and above. The term GVCF is sometimes used simply to describe VCFs that contain a record for every position in the genome (or interval of interest) regardless of whether a variant was detected at that site or not (such as VCFs produced by UnifiedGenotyper with --output_mode EMIT_ALL_SITES). GVCFs produced by HaplotypeCaller 3.x contain additional information that is formatted in a very specific way. Read on to find out more.

General comparison of VCF vs. gVCF

The key difference between a regular VCF and a gVCF is that the gVCF has records for all sites, whether there is a variant call there or not. The goal is to have every site represented in the file in order to do joint analysis of a cohort in subsequent steps. The records in a gVCF include an accurate estimation of how confident we are in the determination that the sites are homozygous-reference or not. This estimation is generated by the HaplotypeCaller's built-in reference model.

Note that some other tools (including the GATK's own UnifiedGenotyper) may output an all-sites VCF that looks superficially like the BP_RESOLUTION gVCFs produced by HaplotypeCaller, but they do not provide an accurate estimate of reference confidence, and therefore cannot be used in joint genotyping analyses.

The two types of gVCFs

As you can see in the figure above, there are two options you can use with -ERC: GVCF and BP_RESOLUTION. With BP_RESOLUTION, you get a gVCF with an individual record at every site: either a variant record, or a non-variant record. With GVCF, you get a gVCF with individual variant records for variant sites, but the non-variant sites are grouped together into non-variant block records that represent intervals of sites for which the genotype quality (GQ) is within a certain range or band. The GQ ranges are defined in the ##GVCFBlock line of the gVCF header. The purpose of the blocks (also called banding) is to keep file size down, and there is no downside for the downstream analysis, so we do recommend using the -GVCF option.

Example gVCF file

This is a banded gVCF produced by HaplotypeCaller with the -GVCF option.


As you can see in the first line, the basic file format is a valid version 4.1 VCF. See also the ##GVCFBlock lines (after the ##FORMAT lines) which indicate the GQ ranges used for banding, and the definition of the MIN_DP annotation in the ##FORMAT lines.

##ALT=<ID=NON_REF,Description="Represents any possible alternative allele at this location">
##FILTER=<ID=LowQual,Description="Low quality">
##FORMAT=<ID=AD,Number=.,Type=Integer,Description="Allelic depths for the ref and alt alleles in the order listed">
##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Approximate read depth (reads with MQ=255 or with bad mates are filtered)">
##FORMAT=<ID=GQ,Number=1,Type=Integer,Description="Genotype Quality">
##FORMAT=<ID=MIN_DP,Number=1,Type=Integer,Description="Minimum DP observed within the GVCF block">
##FORMAT=<ID=PL,Number=G,Type=Integer,Description="Normalized, Phred-scaled likelihoods for genotypes as defined in the VCF specification">
##FORMAT=<ID=SB,Number=4,Type=Integer,Description="Per-sample component statistics which comprise the Fisher's Exact Test to detect strand bias.">
##INFO=<ID=BaseQRankSum,Number=1,Type=Float,Description="Z-score from Wilcoxon rank sum test of Alt Vs. Ref base qualities">
##INFO=<ID=ClippingRankSum,Number=1,Type=Float,Description="Z-score From Wilcoxon rank sum test of Alt vs. Ref number of hard clipped bases">
##INFO=<ID=DP,Number=1,Type=Integer,Description="Approximate read depth; some reads may have been filtered">
##INFO=<ID=DS,Number=0,Type=Flag,Description="Were any of the samples downsampled?">
##INFO=<ID=END,Number=1,Type=Integer,Description="Stop position of the interval">
##INFO=<ID=HaplotypeScore,Number=1,Type=Float,Description="Consistency of the site with at most two segregating haplotypes">
##INFO=<ID=InbreedingCoeff,Number=1,Type=Float,Description="Inbreeding coefficient as estimated from the genotype likelihoods per-sample when compared against the Hardy-Weinberg expectation">
##INFO=<ID=MLEAC,Number=A,Type=Integer,Description="Maximum likelihood expectation (MLE) for the allele counts (not necessarily the same as the AC), for each ALT allele, in the same order as listed">
##INFO=<ID=MLEAF,Number=A,Type=Float,Description="Maximum likelihood expectation (MLE) for the allele frequency (not necessarily the same as the AF), for each ALT allele, in the same order as listed">
##INFO=<ID=MQ,Number=1,Type=Float,Description="RMS Mapping Quality">
##INFO=<ID=MQ0,Number=1,Type=Integer,Description="Total Mapping Quality Zero Reads">
##INFO=<ID=MQRankSum,Number=1,Type=Float,Description="Z-score From Wilcoxon rank sum test of Alt vs. Ref read mapping qualities">
##INFO=<ID=ReadPosRankSum,Number=1,Type=Float,Description="Z-score from Wilcoxon rank sum test of Alt vs. Ref read position bias">


The first thing you'll notice, hopefully, is the <NON_REF> symbolic allele listed in every record's ALT field. This provides us with a way to represent the possibility of having a non-reference allele at this site, and to indicate our confidence either way.

The second thing to look for is the END tag in the INFO field of non-variant block records. This tells you at what position the block ends. For example, the first line is a non-variant block that starts at position 20:10000000 and ends at 20:10000116.

20  10000000    .   T   <NON_REF>   .   .   END=10000116    GT:DP:GQ:MIN_DP:PL  0/0:44:99:38:0,89,1385
20  10000117    .   C   T,<NON_REF> 612.77  .   BaseQRankSum=0.000;ClippingRankSum=-0.411;DP=38;MLEAC=1,0;MLEAF=0.500,0.00;MQ=221.39;MQ0=0;MQRankSum=-2.172;ReadPosRankSum=-0.235   GT:AD:DP:GQ:PL:SB   0/1:17,21,0:38:99:641,0,456,691,519,1210:6,11,11,10
20  10000118    .   T   <NON_REF>   .   .   END=10000210    GT:DP:GQ:MIN_DP:PL  0/0:42:99:38:0,80,1314
20  10000211    .   C   T,<NON_REF> 638.77  .   BaseQRankSum=0.894;ClippingRankSum=-1.927;DP=42;MLEAC=1,0;MLEAF=0.500,0.00;MQ=221.89;MQ0=0;MQRankSum=-1.750;ReadPosRankSum=1.549    GT:AD:DP:GQ:PL:SB   0/1:20,22,0:42:99:667,0,566,728,632,1360:9,11,12,10
20  10000212    .   A   <NON_REF>   .   .   END=10000438    GT:DP:GQ:MIN_DP:PL  0/0:52:99:42:0,99,1403
20  10000439    .   T   G,<NON_REF> 1737.77 .   DP=57;MLEAC=2,0;MLEAF=1.00,0.00;MQ=221.41;MQ0=0 GT:AD:DP:GQ:PL:SB   1/1:0,56,0:56:99:1771,168,0,1771,168,1771:0,0,0,0
20  10000440    .   T   <NON_REF>   .   .   END=10000597    GT:DP:GQ:MIN_DP:PL  0/0:56:99:49:0,120,1800
20  10000598    .   T   A,<NON_REF> 1754.77 .   DP=54;MLEAC=2,0;MLEAF=1.00,0.00;MQ=185.55;MQ0=0 GT:AD:DP:GQ:PL:SB   1/1:0,53,0:53:99:1788,158,0,1788,158,1788:0,0,0,0
20  10000599    .   T   <NON_REF>   .   .   END=10000693    GT:DP:GQ:MIN_DP:PL  0/0:51:99:47:0,120,1800
20  10000694    .   G   A,<NON_REF> 961.77  .   BaseQRankSum=0.736;ClippingRankSum=-0.009;DP=54;MLEAC=1,0;MLEAF=0.500,0.00;MQ=106.92;MQ0=0;MQRankSum=0.482;ReadPosRankSum=1.537 GT:AD:DP:GQ:PL:SB   0/1:21,32,0:53:99:990,0,579,1053,675,1728:9,12,10,22
20  10000695    .   G   <NON_REF>   .   .   END=10000757    GT:DP:GQ:MIN_DP:PL  0/0:48:99:45:0,120,1800
20  10000758    .   T   A,<NON_REF> 1663.77 .   DP=51;MLEAC=2,0;MLEAF=1.00,0.00;MQ=59.32;MQ0=0  GT:AD:DP:GQ:PL:SB   1/1:0,50,0:50:99:1697,149,0,1697,149,1697:0,0,0,0
20  10000759    .   A   <NON_REF>   .   .   END=10001018    GT:DP:GQ:MIN_DP:PL  0/0:40:99:28:0,65,1080
20  10001019    .   T   G,<NON_REF> 93.77   .   BaseQRankSum=0.058;ClippingRankSum=-0.347;DP=26;MLEAC=1,0;MLEAF=0.500,0.00;MQ=29.65;MQ0=0;MQRankSum=-0.925;ReadPosRankSum=0.000 GT:AD:DP:GQ:PL:SB   0/1:19,7,0:26:99:122,0,494,179,515,694:12,7,4,3
20  10001020    .   C   <NON_REF>   .   .   END=10001020    GT:DP:GQ:MIN_DP:PL  0/0:26:72:26:0,72,1080
20  10001021    .   T   <NON_REF>   .   .   END=10001021    GT:DP:GQ:MIN_DP:PL  0/0:25:37:25:0,37,909
20  10001022    .   C   <NON_REF>   .   .   END=10001297    GT:DP:GQ:MIN_DP:PL  0/0:30:87:25:0,72,831
20  10001298    .   T   A,<NON_REF> 1404.77 .   DP=41;MLEAC=2,0;MLEAF=1.00,0.00;MQ=171.56;MQ0=0 GT:AD:DP:GQ:PL:SB   1/1:0,41,0:41:99:1438,123,0,1438,123,1438:0,0,0,0
20  10001299    .   C   <NON_REF>   .   .   END=10001386    GT:DP:GQ:MIN_DP:PL  0/0:43:99:39:0,95,1226
20  10001387    .   C   <NON_REF>   .   .   END=10001418    GT:DP:GQ:MIN_DP:PL  0/0:41:42:39:0,21,315
20  10001419    .   T   <NON_REF>   .   .   END=10001425    GT:DP:GQ:MIN_DP:PL  0/0:45:12:42:0,9,135
20  10001426    .   A   <NON_REF>   .   .   END=10001427    GT:DP:GQ:MIN_DP:PL  0/0:49:0:48:0,0,1282
20  10001428    .   T   <NON_REF>   .   .   END=10001428    GT:DP:GQ:MIN_DP:PL  0/0:49:21:49:0,21,315
20  10001429    .   G   <NON_REF>   .   .   END=10001429    GT:DP:GQ:MIN_DP:PL  0/0:47:18:47:0,18,270
20  10001430    .   G   <NON_REF>   .   .   END=10001431    GT:DP:GQ:MIN_DP:PL  0/0:45:0:44:0,0,1121
20  10001432    .   A   <NON_REF>   .   .   END=10001432    GT:DP:GQ:MIN_DP:PL  0/0:43:18:43:0,18,270
20  10001433    .   T   <NON_REF>   .   .   END=10001433    GT:DP:GQ:MIN_DP:PL  0/0:44:0:44:0,0,1201
20  10001434    .   G   <NON_REF>   .   .   END=10001434    GT:DP:GQ:MIN_DP:PL  0/0:44:18:44:0,18,270
20  10001435    .   A   <NON_REF>   .   .   END=10001435    GT:DP:GQ:MIN_DP:PL  0/0:44:0:44:0,0,1130
20  10001436    .   A   AAGGCT,<NON_REF>    1845.73 .   DP=43;MLEAC=2,0;MLEAF=1.00,0.00;MQ=220.07;MQ0=0 GT:AD:DP:GQ:PL:SB   1/1:0,42,0:42:99:1886,125,0,1888,126,1890:0,0,0,0
20  10001437    .   A   <NON_REF>   .   .   END=10001437    GT:DP:GQ:MIN_DP:PL  0/0:44:0:44:0,0,0

Note that toward the end of this snippet, you see multiple consecutive non-variant block records. These were not merged into a single record because the sites they contain belong to different ranges of GQ (which are defined in the header).

Comments (9)

The HaplotypeCaller is a more recent and sophisticated tool than the UnifiedGenotyper. Its ability to call SNPs is equivalent to that of the UnifiedGenotyper, and its ability to call indels is far superior. We recommend using HaplotypeCaller in all cases, with only a few exceptions:

  • If you want to analyze more than 100 samples at a time (for performance reasons)
  • If you are working with non-diploid organisms (UG can handle different levels of ploidy while HC cannot)
  • If you are working with pooled samples (also due to the HC’s limitation regarding ploidy)

In those cases, we recommend using UnifiedGenotyper instead of HaplotypeCaller.

Comments (49)


Call variants on a diploid genome with the HaplotypeCaller, producing a raw (unfiltered) VCF.


This is meant only for single-sample analysis. To analyze multiple samples, see the Best Practices documentation on joint analysis.


  • TBD


  1. Determine the basic parameters of the analysis
  2. Call variants in your sequence data

1. Determine the basic parameters of the analysis

If you do not specify these parameters yourself, the program will use default values. However we recommend that you set them explicitly because it will help you understand how the results are bounded and how you can modify the program's behavior.

  • Genotyping mode (–genotyping_mode)

This specifies how we want the program to determine the alternate alleles to use for genotyping. In the default DISCOVERY mode, the program will choose the most likely alleles out of those it sees in the data. In GENOTYPE_GIVEN_ALLELES mode, the program will only use the alleles passed in from a VCF file (using the -alleles argument). This is useful if you just want to determine if a sample has a specific genotype of interest and you are not interested in other alleles.

  • Emission confidence threshold (–stand_emit_conf)

This is the minimum confidence threshold (phred-scaled) at which the program should emit sites that appear to be possibly variant.

  • Calling confidence threshold (–stand_call_conf)

This is the minimum confidence threshold (phred-scaled) at which the program should emit variant sites as called. If a site's associated genotype has a confidence score lower than the calling threshold, the program will emit the site as filtered and will annotate it as LowQual. This threshold separates high confidence calls from low confidence calls.

The terms called and filtered are tricky because they can mean different things depending on context. In ordinary language, people often say a site was called if it was emitted as variant. But in the GATK's technical language, saying a site was called means that that site passed the confidence threshold test. For filtered, it's even more confusing, because in ordinary language, when people say that sites were filtered, they usually mean that those sites successfully passed a filtering test. However, in the GATK's technical language, the same phrase (saying that sites were filtered) means that those sites failed the filtering test. In effect, it means that those would be filtered out if the filter was used to actually remove low-confidence calls from the callset, instead of just tagging them. In both cases, both usages are valid depending on the point of view of the person who is reporting the results. So it's always important to check what is the context when interpreting results that include these terms.

2. Call variants in your sequence data


Run the following GATK command:

java -jar GenomeAnalysisTK.jar \ 
    -T HaplotypeCaller \ 
    -R reference.fa \ 
    -I preprocessed_reads.bam \  # can be reduced or not
    -L 20 \ 
    --genotyping_mode DISCOVERY \ 
    -stand_emit_conf 10 \ 
    -stand_call_conf 30 \ 
    -o raw_variants.vcf 

Note: This is an example command. Please look up what the arguments do and see if they fit your analysis before copying this. To see how the -L argument works, you can refer here: http://gatkforums.broadinstitute.org/discussion/4133/when-should-i-use-l-to-pass-in-a-list-of-intervals#latest

Expected Result

This creates a VCF file called raw_variants.vcf, containing all the sites that the HaplotypeCaller evaluated to be potentially variant. Note that this file contains both SNPs and Indels.

Although you now have a nice fresh set of variant calls, the variant discovery stage is not over. The distinctions made by the caller itself between low-confidence calls and the rest is very primitive, and should not be taken as a definitive guide for filtering. The GATK callers are designed to be very lenient in calling variants, so it is extremely important to apply one of the recommended filtering methods (variant recalibration or hard-filtering), in order to move on to downstream analyses with the highest-quality call set possible.

Comments (5)

Until now, HaplotypeCaller was only capable of calling variants in diploid organisms due to some assumptions made in the underlying algorithms. I'm happy to announce that we now have a generalized version that is capable of handling any ploidy you specify at the command line!

This new feature, which we're calling "omniploidy", is technically still under development, but we think it's mature enough for the more adventurous to try out as a beta test ahead of the next official release. We'd especially love to get some feedback from people who work with non-diploids on a regular basis, so we're hoping that some of you microbiologists and assorted plant scientists will take it out for a spin and let us know how it behaves in your hands.

It's available in the latest nightly builds; just use the -ploidy argument to give it a whirl. If you have any questions or feedback, please post a comment on this article in the forum.

Caveat: the downstream tools involved in the new GVCF-based workflow (GenotypeGVCFs and CombineGVCFs) are not yet capable of handling non-diploid calls correctly -- but we're working on it.

Comments (13)

GATK 3.2 was released on July 14, 2014. Highlights are listed below. Read the detailed version history overview here: http://www.broadinstitute.org/gatk/guide/version-history

We also want to take this opportunity to thank super-user Phillip Dexheimer for all of his excellent contributions to the codebase, especially for this release.

Haplotype Caller

  • Various improvements were made to the assembly engine and likelihood calculation, which leads to more accurate genotype likelihoods (and hence better genotypes).
  • Reads are now realigned to the most likely haplotype before being used by the annotations, so AD and DP will now correspond directly to the reads that were used to generate the likelihoods.
  • The caller is now more conservative in low complexity regions, which significantly reduces false positive indels at the expense of a little sensitivity; mostly relevant for whole genome calling.
  • Small performance optimizations to the function to calculate the log of exponentials and to the Smith-Waterman code (thanks to Nigel Delaney).
  • Fixed small bug where indel discovery was inconsistent based on the active-region size.
  • Removed scary warning messages for "VectorPairHMM".
  • Made VECTOR_LOGLESS_CACHING the default implementation for PairHMM.
  • When we subset PLs because alleles are removed during genotyping we now also subset the AD.
  • Fixed bug where reference sample depth was dropped in the DP annotation.

Variant Recalibrator

  • The -mode argument is now required.
  • The plotting script now uses the theme instead of opt functions to work with recent versions of the ggplot2 R library.


  • The plotting script now uses the theme instead of opt functions to work with recent versions of the ggplot2 R library.

Variant Annotator

  • SB tables are created even if the ref or alt columns have no counts (used in the FS and SOR annotations).

Genotype GVCFs

  • Added missing arguments so that now it models more closely what's available in the Haplotype Caller.
  • Fixed recurring error about missing PLs.
  • No longer pulls the headers from all input rods including dbSNP, rather just from the input variants.
  • --includeNonVariantSites should now be working.

Select Variants

  • The dreaded "Invalid JEXL expression detected" error is now a kinder user error.

Indel Realigner

  • Now throws a user error when it encounters reads with I operators greater than the number of read bases.
  • Fixed bug where reads that are all insertions (e.g. 50I) were causing it to fail.


  • Now computes posterior probabilities only for SNP sites with SNP priors (other sites have flat priors applied).
  • Now computes genotype posteriors using likelihoods from all members of the trio.
  • Added annotations for calling potential de novo mutations.
  • Now uses PP tag instead of GP tag because posteriors are Phred-scaled.

Cat Variants

  • Can now process .list files with -V.
  • Can now handle BCF and Block-Compressed VCF files.

Validate Variants

  • Now works with gVCF files.
  • By default, all strict validations are performed; use --validationTypeToExclude to exclude specific tests.


  • Now use '--use_IUPAC_sample sample_name' to specify which sample's genotypes should be used for the IUPAC encoding with multi-sample VCF files.


  • Refactored maven directories and java packages replacing "sting" with "gatk".
  • Extended on-the-fly sample renaming feature to VCFs with the --sample_rename_mapping_file argument.
  • Added a new read transformer that refactors NDN cigar elements to one N element.
  • Now a Tabix index is created for block-compressed output formats.
  • Switched outputRoot in SplitSamFile to an empty string instead of null (thanks to Carlos Barroto).
  • Enabled the AB annotation in the reference model pipeline (thanks to John Wallace).
  • We now check that output files are specified in a writeable location.
  • We now allow blank lines in a (non-BAM) list file.
  • Added legibility improvements to the Progress Meter.
  • Allow for non-tab whitespace in sample names when performing on-the-fly sample-renaming (thanks to Mike McCowan).
  • Made IntervalSharder respect the IntervalMergingRule specified on the command line.
  • Sam, tribble, and variant jars updated to version 1.109.1722; htsjdk updated to version 1.112.1452.
Comments (0)

GATK 3.1 was released on March 18, 2014. Highlights are listed below. Read the detailed version history overview here: http://www.broadinstitute.org/gatk/guide/version-history

Haplotype Caller

  • Added new capabilities to the Haplotype Caller to use hardware-based optimizations. Can be enabled with --pair_hmm_implementation VECTOR_LOGLESS_CACHING. Please see the 3.1 Version Highlights for more details about expected speed ups and some background on the collaboration that made these possible.
  • Fixed bugs in computing the weights of edges in the assembly graph. This was causing bad genotypes to be output when running the Haplotype Caller over multiple samples simultaneously (as opposed to creating gVCFs in the new recommended pipeline, which was working as expected).

Variant Recalibrator

  • Fixed issue where output could be non-deterministic with very large data sets.


  • Fixed several bugs where bad input were causing the tool to crash instead of gracefully exiting with an error message.


  • RandomlySplitVariants can now output splits comprised of more than 2 output files.
  • FastaAlternateReferenceMaker can now output heterozygous sites using IUPAC ambiguity encoding.
  • Picard, Tribble, and Variant jars updated to version 1.109.1722.
Comments (30)

As reported here:

  • http://gatkforums.broadinstitute.org/discussion/2342/unifiedgenotyper-causes-arrayindexoutofboundsexception-3-how-to-fix-it
  • http://gatkforums.broadinstitute.org/discussion/2343/printreads-yet-another-arrayindexoutofboundsexception
  • http://gatkforums.broadinstitute.org/discussion/2345/arrayindexoutofboundsexception-in-haplotypecaller

If you encounter this bug too, please don't post a new question about it. Feel free to comment in this thread to let us know you have also had the same problem. Tell us what version of the GATK you were using and post your command line.

Thank you for your patience while we work to fix this issue.

Update: it seems the cases in the different tools aren't directly related. The PrintReads and HaplotypeCaller cases are now solved; still waiting on test files for the UG case.

No posts found with the requested search criteria.