Unraveling overlapping deletions by agglomerative clustering

Abstract

BackgroundStructural variations in human genomes, such as deletions, play an important role in cancer development. Next-Generation Sequencing technologies have been central in providing ways to detect such variations. Methods like paired-end mapping allow to simultaneously analyze data from several samples in order to, e.g., distinguish tumor from patient specific variations. However, it has been shown that, especially in this setting, there is a need to explicitly take overlapping deletions into consideration. Existing tools have only minor capabilities to call overlapping deletions, unable to unravel complex signals to obtain consistent predictions.ResultWe present a first approach specifically designed to cluster short-read paired-end data into possibly overlapping deletion predictions. The method does not make any assumptions on the composition of the data, such as the number of samples, heterogeneity, polyploidy, etc. Taking paired ends mapped to a reference genome as input, it iteratively merges mappings to clusters based on a similarity score that takes both the putative location and size of a deletion into account.ConclusionWe demonstrate that agglomerative clustering is suitable to predict deletions. Analyzing real data from three samples of a cancer patient, we found putatively overlapping deletions and observed that, as a side-effect, erroneous mappings are mostly identified as singleton clusters. An evaluation on simulated data shows, compared to other methods which can output overlapping clusters, high accuracy in separating overlapping from single deletions.

Diploid Alignments and Haplotyping

“…As the motivation for our approach is complex genome regions, it should be observed that just the detection of variants in such areas is challenging. It can happen that variant predictions overlap such that at some positions a diploid genome is not enough to cover all variants [16]. For the process in Sect.…”

Bioinformatics Research and Applications

“…As the motivation for our approach is complex genome regions, it should be observed that just the detection of variants in such areas is challenging. It can happen that variant predictions overlap such that at some positions a diploid genome is not enough to cover all variants [16]. For the process in Sect.…”

Haploid to diploid alignment for variation calling assessment

“…In the general case, one could be left with overlapping mutations. Then one should construct another haploid genome applying all homozygous mutations as well; it is possible that there are more overlaps because of prediction errors [ 4 ], but since our simulated ground-truth is diploid, generating more predicted genomes is not beneficial in this case.…”