Academic literature on the topic 'Somatic variant calling'

Abstract Motivation: Somatic variant calling typically requires paired tumor-normal tissue samples. Yet, paired normal tissues are not always available in clinical settings or for archival samples. Results: We present SomVarIUS, a computational method for detecting somatic variants using high throughput sequencing data from unpaired tissue samples. We evaluate the performance of the method using genomic data from synthetic and real tumor samples. SomVarIUS identifies somatic variants in exome-seq data of ∼150 × coverage with at least 67.7% precision and 64.6% recall rates, when compared with paired-tissue somatic variant calls in real tumor samples. We demonstrate the utility of SomVarIUS by identifying somatic mutations in formalin-fixed samples, and tracking clonal dynamics of oncogenic mutations in targeted deep sequencing data from pre- and post-treatment leukemia samples. Availability and implementation: SomVarIUS is written in Python 2.7 and available at http://www.sjdlab.org/resources/ Contact: subhajyoti.de@ucdenver.edu Supplementary information: Supplementary data are available at Bioinformatics online.

Abstract Background: The human genome is very heterogeneous on the individual level which challenges interpretation of whole genome sequencing (WGS) data. In order to reduce complexity in tumor genetics WGS of a tumor is performed together with WGS of "normal" tissue from the respective patient (i.e. fingernails, skin biopsy, hair, buccal swaps) which is used as the germline sequence (tumor/matched normal approach, TMNA). This approach allows the extraction of somatic mutations acquired in the tumor through sophisticated algorithms. In routine diagnostics, especially in hematological neoplasms, "normal" tissue representing the germline sequence is usually not available, which prohibits the standard use of somatic tumor/normal variant calling tools. Aims: On the road to implement WGS into routine diagnostics we tested a TMNA in comparison to a tumor/unmatched normal approach (TUNA), where pooled genomic DNA (Promega, Fitchburg, WI) was used instead of a matched normal. Cohorts and Methods: 9 samples from patients with hematological neoplasms (7 AML, 2 ALL) were sequenced at diagnosis on Illumina HiSeqX machines (Illumina, San Diego, CA), along with complete remission samples to serve as matched normals for the TMNA. For comparison, a mixture of genomic DNA from multiple anonymous donors was used as "normal" for the TUNA. Read mapping and somatic variant calling was performed using the tools Isaac3 and Strelka2, respectively. Statistical differences between groups were assessed by two-sided Mann-Whitney tests. Results: The TMNA produced a median of 17,700 somatic variant calls, while the TUNA produced 419,000. This 24-fold disparity is mainly due to residual germline variants missed by the TUNA. A large fraction of TMNA variants (57%) was located in regions of known low confidence variant calling (as defined by the Genome in a Bottle Consortium) and likely contain mostly artifacts. After removing these regions from analysis a median of 7,700 and 331,000 variants remained in the TMNA and TUNA datasets, respectively. In order to eliminate germline variants, the gnomAD population database was queried and any present variants were discarded. As expected, this removed over 95% of all variants from the TUNA dataset, but also 41% from the TMNA dataset. The latter might be attributed to common germline variants falsely being called as somatic by the TMNA and/or somatic mutations occurring at polymorphic sites. After this filtering step a median of 3,770 and 15,500 variants remained in the TMNA and TUNA datasets, respectively. This 4-fold disparity in variant number is most likely caused by rare germline variation remaining in the TUNA dataset. Of the remaining TMNA variants only 65% could be found within the larger TUNA dataset. A major factor governing this observation was variant allele frequency (VAF). Variants that overlapped between both datasets had on average higher VAFs than those unique to the TMNA (p < 2.2x10-16). Further inspection of the VAF distribution among samples revealed a bimodal or nearly bimodal distribution for all samples. All distributions shared a sharp peak centered on a VAF of 10%, which was unexpected given the estimated tumor fractions of the samples predict VAFs of 25% and higher. Variants in this lower part of the distribution (arbitrarily defined as VAFs < 20%) constitute on average 50% of all variants in a TMNA sample, with extremes reaching 95% in 2 samples. These low frequency variants show distinctly lower mapping qualities than variants with VAFs ≥ 20% (p < 2.2x10-16), i.e. they reside in regions of elevated mapping ambiguity which potentially leads to the creation of artefacts. Analyzing the overlap of only the higher VAF variants we find that 97.4% of all TMNA variants can also be found in the TUNA dataset. Conclusions: Comparing tumor samples to matched normal material from the respective patient is the preferred approach for somatic variant calling in WGS data, however even with modern algorithms false positives due to technical artifacts seem to be highly abundant. A deeper understanding of the nature of these artifacts is crucial for developing appropriate filtering schemes and improving variant calling algorithms. In the absence of a matched normal using a TUNA can uncover the vast majority (97.4%) of high-quality variants found in a TMNA, however distinguishing true somatic variants from residual rare germline variation in a TUNA remains a major challenge. Disclosures Hutter: MLL Munich Leukemia Laboratory: Employment. Nadarajah:MLL Munich Leukemia Laboratory: Employment. Meggendorfer:MLL Munich Leukemia Laboratory: Employment. Kern:MLL Munich Leukemia Laboratory: Employment, Equity Ownership. Haferlach:MLL Munich Leukemia Laboratory: Employment, Equity Ownership. Haferlach:MLL Munich Leukemia Laboratory: Employment, Equity Ownership.

e13659 Background: Efficient and accurate identification of somatic variant is important for understanding the formation, progression, and treatment of cancer. It is necessary to conduct manual review by Integrative Genomic Viewer (IGV) in traditional variant calling process. However, the traditional manual is heavy workload when evaluating tumor with a high variant burden. In this study, a new convolutional neural network (CNN) method was created to train models for somatic mutation identification, which was suitable for Panel sequencing platform with different tumor purities. Methods: A total of 1000 tumor samples from next generation sequencing (NGS)-based genetic testing by a College of American Pathologists (CAP) accredited and Clinical Laboratory Improvement Amendments (CLIA) certified laboratory. Through variant calling program, like GATK, the candidate mutation locations were identified and standardized by manual confirmation. For each candidate mutation location, reads of both tumor and control tissue were extracted. A 2-dimensional feature matrix M of size (2k+1) * 32 in each candidate base was created. The rows of 2k+1 represented the length of candidate region, and the 32 columns included the reads coverage frequency, mapping quality messages, and genome local scores of different tumor and control tissues. CNN model, which includes nine convolutional layers structured by Temporal Convolutional Networks (TCN) but with a different structure to adapt to the proposed input matrix, was used for training. The training data set including manually validated sequence data was used as benchmark test, and optimized by Stochastic Gradient Descent (SGD) optimizer with a learning rate of 0.01 was used for training. Results: The validation data set included 15 mixed samples which were composed of different proportions of known cell lines and real mixed blood samples. The pooled DNA contained 2,359 somatic variants, with expected variant allele frequencies ranged from 3% to 97% in each pool. The overall sensitivity and positive predictive value (PPV) of single nucleotide variants (SNVs) were 99.3% and 99.8%, respectively. Conclusions: A novel and sensitive computational tool for somatic variation detection in DNA Panel sequencing was developed. Our result showed that the deep learning CNN model could call variant in Panel sequencing data.