BAM Files Research Articles

DEVELOPMENT OF A DATA PORTAL FOR AGGREGATION AND ANALYSIS OF GENOMICS DATA IN FAMILIAL PLATELET DISORDER WITH PREDISPOSITION TO MYELOID MALIGNANCY – THE RUNX1.DB

Over 100 different families have been reported with damaging germline mutations or deletions in RUNX1, making it a clinically significant gene in FPD-MM. In large families, the diagnosis of haematological malignancy (HM) shows variable penetrance among members with the same mutation; some carriers do not develop HM. The cause of this heterogeneity is not yet understood but it is likely attributable to additional somatic genetic changes in these germline RUNX1 carriers. To better understand the genetic landscape surrounding RUNX1 germline predisposition to HM we have developed an interactive online database, the RUNX1 database ( https://runx1db.runx1.com ). This database is designed for the sharing and curation of genetic data acquired from individuals with germline RUNX1 mutation both pre- and post-HM progression. Currently we have collected 170 datasets from 11 countries. Raw NGS data files were transferred, preferentially as FASTQ files, and run through a standardised bioinformatics pipeline generating VCF and BAM files. These files were uploaded onto the RUNX1.db where the integrated VariantGrid, genomics analysis software, can be used to provide automated annotation of variants within the VCF files. Using the RUNX1db to curate variants from previously published and new datasets we have been able to confirm published somatic and germline mutations, and importantly, extend on these findings by identifying additional somatic mutations of clinical significance. Interrogating this dataset as a single cohort we have identified somatic genetic signatures in pre-leukemic samples prior progression to AML, including recurrent mutations in several clonal haematopoiesis genes including BCOR. This RUNX1 worldwide genomics cohort will lead to powerful insights in the development of precision-based medicine diagnosis, risk assessment, monitoring and therapeutic intervention.

24. PREIMPLANTATION GENETIC TESTING FOR ANEUPLOIDIES WITH NGS ON MINISEQ. A RELIABLE ALTERNATIVE FOR SMALL LABORATORIES

Introduction The demand for Preimplantation Genetic Testing for aneuploidy (PGT-A) used in IVF, is constantly increasing to improve pregnancy outcomes by transferring euploid embryos. The introduction of NGS has led to rapid transition of PGT-A from array platforms to low depth high-throughput whole genome sequencing which is now implemented in many laboratories worldwide. However, for small laboratories, the transition from array platforms to NGS should also take into consideration cost effectiveness. The aim of this study was to achieve the transition from array-CGH to NGS using the VeriSeqTM PGS (Illumina Inc.), on the MiniSeq, which is the existing platform of the Department. Consequently, the current study compares PGT-A results between array-CGH and NGS and also directly compares NGS results from MiniSeq and MiSeq, the platform for which VeriSeqTM PGS is already validated. Materials and Methods In total 110 amplified biopsy samples were included and blindly analyzed in this study which was divided into three phases. The first included 30 biopsies of day-3(13) and day-5(17) embryos that were previously analyzed using array-CGH. These samples were sequenced on both MiSeqDx and MiniSeq platforms (Illumina Inc.). For the second phase, 80 amplified biopsy samples (day-5), with segmental, mosaic or full aneuploidies were run on the MiniSeq platform with paired-end (2 × 36) sequencing using the MiniSeq High Output Reagent Kit (75-cycles) (Illumina Inc.). The last phase was to detect the sensitivity for mosaicism by mixing serial diluted samples with different chromosomal losses/gains to give 50%, 40%, 30% and 20% aneuploidy. Library preparations on all phases were carried out using VeriSeqTM PGS (Illumina Inc.) following manufacturer's recommendations. For the MiniSeq, the secondary analysis of alignment was performed on-instrument with BWA-MEM to produce aligned reads. An in-house pipeline was implemented to adjust the headers within bam files in order to facilitate the downstream analysis of samples on Bluefuse Multi v4.4. Results Comparative results on both platforms MiniSeq and MiSeq demonstrated 100% concordance, placing our in house pipeline using MiniSeq platform as a cost effective alternative approach for PGT-A, especially for small laboratories. The two different techniques array-CGH and NGS showed concordant results in 105/107 (98%) of the samples tested and overall concordance rate of 99.8% (2410/2415) per analyzed chromosome. All differences detected were attributed to mosaicism. Three day-3 embryos were excluded from the analysis as they failed to pass the VeriSeqTM PGS quality criteria. NGS successfully detected all segmental, mosaic or full aneuploidies (smallest 15Mb) depicting superior profiles than array-CGH due to the broader dynamic range. The sensitivity of the NGS platform for mosaicism detection was defined to be 20%. Conclusion Our validation study result demonstrates that VeriSeq is fully compatible with the MiniSeq platform and in full concordance with results obtained from the MiSeq. The results showed that PGT-A with NGS can identify more precisely chromosomal mosaicism and segmental aneuploidy than the aCGH platform. With the cost of sequencing gradually decreasing, the cost of PGT-A will make the service accessible to all fertility patients in order to enhance their chances of successful pregnancy.

54. IMPLEMENTATION OF FIRST NIPT SERVICE LABORATORY IN AN INDIAN HOSPITAL

Introduction Non-Invasive Prenatal Testing (NIPT) for common fetal aneuploidies is rapidly increasing in India for both naturally conceived pregnancies and pregnancies conceived by assisted reproductive technologies (ART). This study aims to establish the very first next generation sequencing (NGS) laboratory for NIPT in an Indian hospital using the Ion Torrent platform to provide a complete in-house solution for pregnant women. Materials and Methods In brief, all plasma DNA samples (both control and test groups) were converted to genomic DNA libraries using Fragment Library Kit (Thermo Fisher Scientific, Cat. No. 4471252) with Ion Xpress Barcode-Adapters mix (Thermo Fisher Scientific, Cat. No. 4471250) using our laboratory developed protocol. The libraries were then quantified using Qubit dsDNA HS (high sensitivity) Assay Kit, (Thermo Fisher Scientific) and diluted to equimolar amounts before pooling. The Ion Chef System (Thermo Fisher Scientific), as part of the workflow, was used for automated template preparation and loading the sequencing chip. Thereafter, low-pass whole-genome sequencing was done using the ion semiconductor sequencer Ion S5 System (Thermo Fisher Scientific,) to achieve at least 5 million sequence reads. In the dry-lab, the sequenced reads were aligned to the reference human genome (UCSC hg19) locally and their BAM files were generated. The euploid BAM files from the training set of samples were assigned to NIPT reference group and their aligned reads were partitioned into 50,000 bp bins. The number of reads in each bin becomes the unit for comparison with the test samples. GC corrections were performed based on the average coverage of genomic regions having a similar GC-content to correct the coverage differences between bins having a different GC percentage. In this validation study, the standard Z-score prediction models returned a value between -3 and +3 for all the chromosomes, indicating that there is no trisomy present. The scores for aneuploidy samples all showed a value above 4. The z-score refers to the number of standard deviations from the mean of a reference data set and in all molecular counting approaches, a z-score of more than 4 is considered a positive result. Results The algorithm was blindly tested on a cohort of 100 known samples, of which the 12 were T21, eight were T18, four were T13 and 76 were from euploid pregnancies. Our algorithm correctly called all euploid and aneuploid cases. For all aneuploidies, z scores were above a value of 4. Interestingly, accurate detection of aneuploidy was also noted at foetal fraction as low as 3.2%. Therefore, based on our findings, our algorithm is expected to be sensitive and specific for up to 99.9% of pregnancies with one of the three common trisomies. DISCUSSION Based on the high performance of our NIPT test in validation studies, testing has commenced at Lifeline Hospital, including pregnancies conceived by ART. Clinical performance is currently under evaluation to derive positive predictive value (PPV) and negative predictive value (NPV) for each trisomy.

12. EMBRYOS FROM THE SAME COHORT SHOW HIGH VARIABILITY IN THE MTDNA LEVELS

Introduction Mitochondria play a vital role in cell function. Recently, there has been an increasing research on mtDNA as biomarker of embryo implantation. Reports showed that high levels of mtDNA in blastocyst are associated with low implantation potential. Based on the available evidence, it is unclear whether high levels of mtDNA is a constant feature in some group of women or it depends on each embryo, showing variability among sibling embryos from the same cohort. The aim of this study was to investigate the variability in the levels of mtDNA among sibling euploid embryos from the same cohort. Material and Methods A prospective non-selection study was performed from January 2018 to December 2018. We included 249 blastocysts biopsies from 89 couples who attended our clinic for PGT-A with oocyte donation and with at least two of their biopsied embryos diagnosed as euploid non-mosaic. The aneuploid testing was performed by NGS-Veriseq® (Illumina). The mtDNA analysis was performed once the embryo diagnosis was known using MitoSeek. Sequencing reads mapping to the mtDNA genome were extracted from bam files. The relative measure of mtDNA copy number was calculated by dividing the mtDNA reads by the nuclearDNA and subjected to mathematical correction factor according to the embryo genome. Given that the mtDNA content is dependent upon the number of cell division that preceded the biopsy, the mtDNA content was also corrected by a factor according to the day when the embryos were biopsied. Statistical analysis was performed using SPSSv20.0. Results In order to avoid confusion related to the oocyte quality we included only egg donation cycles, therefore the average maternal age at time of oocyte retrieval was 26± 4 years (range: 19.0–33.0). Overall, the mean average of mtDNA per genome was 0.0010±0.0008 (range: 0.0002-0.0039). We analysed the mtDNA within an embryo cohort. To evaluate the extent of the variability in the mtDNA among embryos from the same cohort we calculate the coefficient of variation, being the average 28.77%. Furthermore, each embryo was classified according to the mtDNA 10th-percentile which belongs. The mean difference in the 10th-percentile between the embryo with the higher and the lower mtDNA within a cohort was 30. This difference was not related with the age of the egg donor neither the male factor (p>0.05). Only 13.5% of patients showed no difference in the mtDNA content within the embryos from the same cohort. Conclusion Our results suggest high variability in the mtDNA content within euploid embryos from the same cohort. MtDNA could be used to predict reproductive potential amongst chromosomally normal embryos however the utility of quantifying mtDNA levels depend on the existence of variability among embryos from the same cohort. This study reports that mtDNA content is not a marker of prognosis amongst patients because it is variable among the embryos from the same cohort, thus mtDNA could be used as an additional marker of embryo competence and a tool for prioritize the euploid embryo to be transferred.

Abstract 5116: Benchmarking the HLA typing performance of three HLA assay and two software

Abstract Background:The adaptive immune response intrinsically depends on hyper variable human leukocyte antigen (HLA) genes. Correct HLA typing is crucial for successful donor-patient matching in organ transplantation and recently is also an emerging biomarker for immune checkpoint inhibitor. With the great progress on next generation sequencing (NGS) technology, sequencing accuracy and throughput have increased while the cost for data have decreased faster than Moore's Law, thus HLA tying by NGS has been widely used from laboratory to clinic. Method:We tested three NGS assay for HLA typing and two recently published HLA typing software from 24 samples. Three assays involve two target capture methods (WES and HLA target sequencing) and one amplification sequencing method. Two software involve Xhla and HLAHD, respectively. First, we compared 3 assay performance based on PE150 read length and xHLA software by both 4 alleles and 8 alleles. Second, we compared the performance of two widespread HLA typing software xHLA and HLA-HD according to different assay type, read depth, read length and speed. Results:First of all, for PE150 datasets on three assay, the accuracy rate of HLAHD is all 100%, while xhla in WES is 99%, amplicom 98.6%, target HLA sequencing 100%. On reads length evaluation, the accuracy of HLAHD decreases gradually from PE150 to PE100 and PE75, and the capture method is more sensitive to reads length than the amplification assay, while Xhla is very robust on read length. On coverage evaluation, when the amplicom sequencing data was down-sampled 500X to 10X, the accuracy of HLAHHD and XHLA is both 100% in 50X and above and decrease gradually when less than 50X. In terms of time, with the increase in sequencing depth, running time for both software increase. HLAHD prediction from fastq file is longer than that from bam file, especially when the panel data is relatively large, the time will increase exponentially. Conclusions:Our benchmarking dataset show that current HLA assay and computational methods for HLA typing generally provide satisfactory results, while different assay type matched with different software, read depth and read length do affect the performance. Citation Format: Rui Meng, Tao Zhu, Yu Gong, Yanan Chen, Hua Dong, Ye Yizhou, Fugen Li, Ping Liu, Minya Yao. Benchmarking the HLA typing performance of three HLA assay and two software [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 5116.

Abstract 2480: MutSpliceDB: A database of splice sites mutations effects

Abstract Splice site mutations are one of the well-known classes of genetic alterations playing an important role in biology. Splice site mutations in cancer are most frequently observed as inactivating alterations in tumor suppressor genes (for example, TP53 or RB1) and to a lesser degree as activating alterations in oncogenes (for example MET). Splice site mutations may lead to alterations in mRNA transcripts, causing for example exon(s) inclusion/exclusion or, intron retention. Interpreting the consequences of a specific splice site mutation is not straightforward, especially if the mutation is located outside of the canonical splice sites. Accurate interpretation of the impact a splice site mutation has can further our understanding of biology, influence patient treatment, and in case of germline splice site mutations, may also have relevance to familial disease predisposition. To facilitate the interpretation of splice site mutation effects, we developed MutSpliceDB: https://brb.nci.nih.gov/splicing a public resource of splice sites mutations effects, documenting mutation effect(s) on splicing based on manually reviewed RNA-seq BAM files from sample(s) with particular splice site mutations. Citation Format: Alida Palmisano, Suleyman Vural, Yingdong Zhao, Dmitriy Sonkin. MutSpliceDB: A database of splice sites mutations effects [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 2480.

Abstract 2487: Application of RegTools to TCGA samples for the identification of tumor-specific splice variants

Abstract The interpretation of variants in cancer is frequently focused on direct protein coding alterations. However, this analysis strategy excludes somatic mutations in non-coding regions of the genome and even exonic mutations may have unidentified non-coding consequences. To address this issue, we created RegTools (Feng et al. 2018), a free, open-source software package that integrates analysis of somatic variant calls from genomic data with splice junctions extracted from transcriptomic data in order to efficiently identify variants that may cause aberrant splicing in tumors. Initially, this software was developed using 188 cancer samples from 4 cancer types and enabled us to identify important non-coding variants which would otherwise be undervalued or perhaps even completely ignored by traditional methods and annotators. Here, we present the application of RegTools to over 10,000 TCGA samples in an effort to identify somatic variants that are associated with alternative splicing patterns within these tumors. In order to perform this analysis, we obtained all tumor samples that had somatic variant VCF files and RNA-seq bam files from the Genomic Data Commons (GDC). Included within these samples are more than 700 solid tissue normal samples, which will be used as a panel of normals in order to filter out variants and splicing events that are not tumor specific. We plan to present on the significance of our findings when analyzing these samples within their respective tumor type and across all tumor types. We also plan to present on whether these aberrant splicing events that result from somatic mutations might interfere with predicted neoantigens pVAC-Seq (Hundal et al. 2016). RegTools is freely available and open source (www.regtools.org). Citation Format: Kelsy C. Cotto, Yang-Yang Feng, Zachary L. Skidmore, Obi L. Griffith, Malachi Griffith. Application of RegTools to TCGA samples for the identification of tumor-specific splice variants [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 2487.

Abstract 1660: Identification of allelic imbalance utilizing heterozygous genotype allele frequencies and intensities

Abstract Allelic imbalance (AI) events, such as amplification, deletion or copy-neutral loss-of-heterozygosity (cn-LOH) can result in the activations of oncogenes or inactivations of tumor suppressor genes that are critical to the process of carcinogenesis and metastasis. However, tumor samples often have low tumor-cellularity and small subclones that require sensitive and robust algorithm for AI detections. To overcome this challenge, our lab has developed an original software application hapLOH (Vattathil and Scheet, 2013) that identifies AI with high sensitivity in SNP array data by utilizing heterozygous genotype allele frequencies in SNP array data using a Hidden Markov Model. Motivated by the plethora of next-generation sequencing (NGS) data, we carried this logic and functionality and developed hapLOHseq (San Lucas et al, 2016) for detection of subtle AI in NGS data using reference and alternate allele read depths and a set of haplotype estimates based on 1000 genome data. Recently, we sought to improve AI detection and provide support for different data types by developing haploh-cn. Haploh-cn extends our previous tools by supporting both SNP array and NGS data, and by incorporating intensity data and Log R ratios (LRR) into the underlying Hidden Markov Model for SNP array data, and incorporating sequencing depth for NGS data. In order to evaluate haploh-cn’s performances, we have downloaded The Cancer Genome Atlas Lung Adenocarcinoma whole exome sequencing data and corresponding Affymetrix SNP6 array data. Then, we simulated tumor cellularity by performing in-silico dilution of NGS and SNP array data. For NGS data, we down sampled tumor sequencing BAM files and mixed in matched-normal sequencing reads. For SNP6 array data, we combined the adjusted B allele frequencies of heterozygous sites in the tumor sample and matched-normal sample. Using the AI events detected in the undiluted tumor samples as the gold standard, we assessed the performance of haploh-cn in the computational diluted samples. Our results showed that haploh-cn had high sensitivity without sacrificing specificity in diluted tumor samples. In summary, haploh-cn is a robust and powerful method for profiling subtle AI in NGS and SNP array data. Citation Format: Kyle Chang, Francis A. San Lucas, Zuhal Ozcan, Smruthy Sivakumar, Yasminka A. Jakubek, Richard G. Fowler, Paul Scheet. Identification of allelic imbalance utilizing heterozygous genotype allele frequencies and intensities [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 1660.

Abstract 2483: An integrated bioinformatics pipeline for profiling cancer-immune interaction from whole exome sequencing of pembrolizumab clinical samples

Abstract Purpose We built an integrated analytic pipeline to robustly and comprehensively profile molecular features of pembrolizumab-treated tumors using whole exome sequencing (WES) data from clinical trials across different indications. Experimental design We implemented a computational framework for analysis of WES data generated by different sequencing vendors for 1467 samples from 8 major tumor types. Reads were first processed by BWA-MEM, Picard, and GATK (v2) to generate analysis-ready BAM files. Quality controls (QC) before downstream analysis included Y/X reads ratio for matched normal sample as patient sex prediction and tumor/normal concordance and contamination estimation by Conpair. Key molecular features were evaluated, including tumor mutation burden (TMB) by MuTect (v1) and VEP; HLA-I typing by OptiType; neoantigen burden by NetMHC (v3.4); mutation signature by deconstructSigs; allele-specific copy number by VarScan2 and Sequenza; clonality by PyClone; presence of oncogenic viruses (eg, EBV, HBV, HPV); and LOH score indicating homologous recombination deficiency (HRD-LOH). Results Concordance rate between predicted and clinical sex was 1446/1464 (98.8%), and 1420/1467 (96.8%) samples passed tumor-normal alignment QC. Highest TMB (median [range]) was detected in melanoma (245 [2-6246]) and urothelial carcinoma (UC) (124 [4-1579]), with lowest TMB in metastatic castration-resistant prostate cancer (52 [1-6143]) and ovarian cancer (OvCa) (49.5 [8-272]). The correlation of median TMB across 7 cancer types in our data and TCGA was Spearman R = 0.957. Within indications, there was no difference in TMB distribution for sequencing data originating from different sequencing vendors and TCGA data, which demonstrated concordance across data sets and robust TMB calling by our integrated pipeline. Neoantigen burden strongly correlated with TMB (N = 1420; Spearman R = 0.890). HPV was detected in 20/129 (15.5%) head and neck squamous cell carcinomas (HNSCC) and 4/6 (66.7%) anal cancers; EBV was detected in 8/129 (6.2%) HNSCC and 18/318 (5.7%) gastric cancers (GC). The dominant mutation signatures by disease included APOBEC for UC (135/236) and HNSCC (22/122), alcohol for HCC (19/35), HRD for OvCa (12/64) and triple-negative breast cancer (51/175), UV exposure for melanoma (145/176), and dMMR for GC (84/287) and CRPC (16/155). Samples with deleterious BRCA mutations showed significantly higher HRD-LOH score (N = 1420; AUROC = 0.61 [95% CI, 0.53-0.69]) and HRD mutational signature (N = 1316; DOR = 6.4 [95% CI, 3.7-11.1]). Conclusion We assembled heterogeneous computational modules into an integrated pipeline to reliably profile diverse molecular features from WES data of nearly 1500 clinical samples across different tumor types. These data serve as a foundation for translational research efforts supporting pembrolizumab development. Citation Format: Xiaoqiao Liu, Xinwei Sher, Hongchao Lu, Jun Zhuang, Weilong Zhao, Andrew Albright, Cinthia Umemoto, Christen Wudarski, Maureen Lane, Mark Ayers, Andrea L. Webber, Sandra C. Souza, Ping Qiu, Diane Levitan, Jennifer Cho, Deepti Aurora-Garg, Matthew Marton, Alexandra Snyder, Michael Morrissey, Andrey Loboda, Ronghua Chen, Razvan Cristescu. An integrated bioinformatics pipeline for profiling cancer-immune interaction from whole exome sequencing of pembrolizumab clinical samples [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 2483.

BAMixChecker: an automated checkup tool for matched sample pairs in NGS cohort.

SummaryMislabeling in the process of next generation sequencing is a frequent problem that can cause an entire genomic analysis to fail, and a regular cohort-level checkup is needed to ensure that it has not occurred. We developed a new, automated tool (BAMixChecker) that accurately detects sample mismatches from a given BAM file cohort with minimal user intervention. BAMixChecker uses a flexible, data-specific set of single-nucleotide polymorphisms and detects orphan (unpaired) and swapped (mispaired) samples based on genotype-concordance score and entropy-based file name analysis. BAMixChecker shows ∼100% accuracy in real WES, RNA-Seq and targeted sequencing data cohorts, even for small panels (<50 genes). BAMixChecker provides an HTML-style report that graphically outlines the sample matching status in tables and heatmaps, with which users can quickly inspect any mismatch events.Availability and implementationBAMixChecker is available at https://github.com/heinc1010/BAMixCheckerSupplementary information Supplementary data are available at Bioinformatics online.

A comparison of three programming languages for a full-fledged next-generation sequencing tool

BackgroundelPrep is an established multi-threaded framework for preparing SAM and BAM files in sequencing pipelines. To achieve good performance, its software architecture makes only a single pass through a SAM/BAM file for multiple preparation steps, and keeps sequencing data as much as possible in main memory. Similar to other SAM/BAM tools, management of heap memory is a complex task in elPrep, and it became a serious productivity bottleneck in its original implementation language during recent further development of elPrep. We therefore investigated three alternative programming languages: Go and Java using a concurrent, parallel garbage collector on the one hand, and C++17 using reference counting on the other hand for handling large amounts of heap objects. We reimplemented elPrep in all three languages and benchmarked their runtime performance and memory use.ResultsThe Go implementation performs best, yielding the best balance between runtime performance and memory use. While the Java benchmarks report a somewhat faster runtime than the Go benchmarks, the memory use of the Java runs is significantly higher. The C++17 benchmarks run significantly slower than both Go and Java, while using somewhat more memory than the Go runs. Our analysis shows that concurrent, parallel garbage collection is better at managing a large heap of objects than reference counting in our case.ConclusionsBased on our benchmark results, we selected Go as our new implementation language for elPrep, and recommend considering Go as a good candidate for developing other bioinformatics tools for processing SAM/BAM data as well.

Correlation of HERVK with efficacy on anti-PD-1/PD-L1 therapy in a pilot metagenomics study on Chinese advanced lung cancer patients.

e14046 Background: Patient selection for PD-1 inhibitors in NSCLC is still based on imperfect screening biomarkers, including PD-L1 expression and tumor mutational burden(TMB).There are several reports show certain bacteria or virus infections are correlated with efficacy and prognosis on PD-1 inhibitors treatment. In order to explore more potential biomarkers for anti-PD-1 therapy, we investigate whether we can discover microorganisms in Chinese lung cancer patients and whether certain microorganism are correlated with efficacy or prognosis on anti-PD-1 therapy using metagenomics technology. Methods: In this study, 30 advanced lung cancer patient were enrolled, including 15 cases of adenocarcinoma, 12 cases of squamous carcinoma, 2 cases of Small cell lung cancer. 10 patients received PD-1/PD-L1 inhibitors as first-line therapy and other patients as second-line therapy. Among these 30 patients, 15 PR,1 CR, 10 PD, 4 SD according to RECIST 1.1. Plasma cell-free DNA (cfDNA) were collected from these patients prior to anti-PD-1 treatment and whole-genome sequencing of cfDNA at average coverage 5X was performed on Illumina NovaSeq in 2 ×150bp length model. A bioinformatics pipeline composed of 4 tools: BWA/Sambama/Kraken2/Bracken, was constructed to analyze the sequencing data. Non-human derived reads were extracted by sambama from the bam file that generated by BWA. Bacteria/Virus abundance in each sample was estimated through Bracken, followed by bac/vir database alignment in Kraken2. Results: In our pilot study, 4 patients were selected for metagenomics sequencing(1CR, 1PR, 2PD). Interestingly, over 100 different microbial sequence were detected in every patient. Compare microorganism in CR+PR patients to PD patients, Human endogenous retrovirus K (HERVK) were quite abundant in CR/PR patients but not in PD patients: 5.6 times more abundant in CR/PR patients comparing to PD patients. (HERVK abundance: CR:PR:PD = 1.20% : 0.32% : (0.17%, 0.10%)). HERVK is a family of human endogenous retroviruses associated with certain malignant tumors, studies have indicated that HERVK may trigger an antigen specific immune response for certain cancer patients. Conclusions: Metagenomics sequencing on plasma cfDNA could be a potential biomarker discovery approach to identify prognostic biomarkers for anti-PD-1 therapy. HERVK are correlated with efficacy on anti-PD-1 treatment in our pilot study, however more thorough investigation need to be conducted to confirm whether HERVK can be served as efficacy biomarkers for anti-PD-1.

Targeted resequencing using the MinION long read sequencing platform

MinION nanopore DNA sequencing is an emerging paradigm in portable, long read, economic sequencing technology. Using a targeted resequencing approach of the rat osteopontin (Spp1) region in the stroke‐prone spontaneously hypertensive (SHRSPGla) and Wistar Kyoto (WKYGla) rats. We aim to develop and validate a variant calling pipeline based on long reads produced by the MinION platform and to corroborate with published SHRSPGla and WKYGla promotor variants using the Rat Genome Database (RGD) (Rnor6.0).Using the Phusion high fidelity DNA polymerase, seven PCR amplifications of defined length were amplified across a 15kb region of the rat Spp1 promoter and gene from liver samples from two individual SHRSPGla and WKYGla animals. The library for MinION sequencing was prepared using Oxford Nanopore Ligation Sequencing Kit and Native Barcoding Expansion 1–12. Purified PCR amplicons were repaired and end‐prepped with dA‐tailing. Barcodes and adapters were ligated to the prepared DNA fragments before loading to the MinION flow cell for sequencing. Two Oxford Nanopore SpotON Flow cells were primed and duplicate prepared samples were each loaded onto flow cells on two separate MinION. The MinION raw data was analysed in real time during sequencing by cloud‐based EPI2ME software (Metrichor). This enabled base calling of raw data to produce DNA sequence from output measurements of current, converting raw fast5 files to base called fastq format. Reads were separated by barcode to identify reads by strain and animal. Quality control metrics for each sample were obtained from the BAM files produced by Samtools Mpileup performed on the merged MinION fastq files mapping to the Spp1 region (Chromosome 14, 6,675,281–6,690,258). A BWA/Bcftools pipeline was used to call variants (Phred&gt;20). Variant calls from the SHRSPGla/WKYGla strains using our MinION pipeline were cross‐referenced with previously published Illumina BWA/GATK calls on RGD for this region (Rnor6.0).The MinION platform produced a mean read count of 1.49×104(± 6.21×103) per sample, with a mean mapping percentage of 49.9% (±10.1). The mean read length was 1130 bases (±132) with a mean quality of base calls of 12.83 (±0.16). Mean depth of coverage of each individual sample was 610x(±180). RGD describes 38 variants in this region of the osteopontin promotor. Across two sequencing runs, 27 of the 38 variants listed on the Rat Genome Database were called and identified several potential novel variants in the region.In conclusion, we confirm the MinION platform can be used as a targeted resequencing platform and variant calls can be made from the long reads produced, although read accuracy is dependent on read coverage.Support or Funding InformationBritish Heart Foundation and Stratified Medicine Scotland Innovation Centre is supported by the Scottish Funding Council, Scottish Enterprise and the Chief Scientist Office of ScotlandThis abstract is from the Experimental Biology 2019 Meeting. There is no full text article associated with this abstract published in The FASEB Journal.

Abstract P4-04-01: The landscape of somatic genetic alterations in breast cancers from CHEK2 germline mutation carriers

Abstract Introduction: Checkpoint kinase 2 (CHEK2) is a tumor suppressor gene, which regulates cell cycle in response to DNA damage response. Selected CHEK2 germline mutations have been shown to confer an increased risk of breast cancer development. Multiple founder mutations in CHEK2 have been identified, and meta analyses have shown that CHEK2 truncating variants confer a higher breast cancer risk than missense variants. Here, we assessed the phenotype and repertoire of genetic alterations of breast cancers from 33 patients with CHEK2 pathogenic germline variants. Materials and methods: We performed targeted capture massively parallel sequencing (≥410 genes) of tumor and normal samples from 13 patients with CHEK2 pathogenic germline variants, and retrieved whole exome sequencing (WES) data (BAM files) of tumor and normal samples from 20 patients with CHEK2 germline pathogenic variants included in the TCGA breast cancer study. In addition, we retrieved WES data of BRCA1, BRCA2 and ATM associated breast cancers from TCGA and Weigelt et al. (JNCI 2018). Somatic mutations, copy number alterations, mutational signatures and large-scale transitions (LSTs) were defined using state-of-the-art bioinformatics algorithms. Results: Of the 33 CHEK2-associated breast cancers included in this study, 21 had missense and 12 had loss-of-function (LoF) germline mutations, and 81% were ER-positive and 12% HER2-positive. CHEK2-associated breast cancers statistically significantly less frequently displayed an ER-negative/HER2-negative phenotype (0%) than BRCA1- (80%) or BRCA2-associated (33%) breast cancers (BRCA1, p&amp;lt;0.0001 for both comparisons), but were similar to ATM-associated breast cancers. Biallelic inactivation of CHEK2 through loss of heterozygosity (LOH) of the wild-type allele was present in 17 of 33 samples (52%). LOH of the CHEK2 wild-type allele was significantly more frequent in tumors with LOF mutations than in those with missense mutations (78% vs 36%, respectively; p=0.0394). PIK3CA (36%) and GATA3 (33%) were the two most recurrently mutated genes in these samples. TP53 somatic mutations were detected in five cases, four of which harbored missense CHEK2 germline mutations. Unlike BRCA1- and BRCA2-associated breast cancers, but akin to ATM-associated breast cancers, CHEK2-associated breast cancers lacked the mutational signature associated with homologous recombination (HR) DNA repair defects (i.e. signature 3) and only five cases displayed high LST scores. Conclusion:CHEK2-associated breast cancers are phenotypically and genetically distinct from BRCA1- and BRCA2-associated breast cancers, but similar to ATM-associated breast cancers. Akin to ATM-associated breast cancers, CHEK2-associated breast cancers are preferentially ER-positive, lack genomics features consistent with defective HR, and have a repertoire of somatic genetic alterations similar to those of non-BRCA1/2 ER-positive breast cancers. Our results suggest that either CHEK2 germline mutations contribute to an increased risk of breast cancer independently of the HR DNA repair defects or that the mutational signatures caused by CHEK2 pathogenic germline mutations differ from those caused by pathogenic germline mutations affecting bona fide HR-related genes (e.g. BRCA1, BRCA2 and PALB2). Citation Format: Kumar R, Pei X, Selenica P, Wen HY, Powell S, Robson M, Riaz N, Reis-Filho JS, Weigelt B, Mandelker D. The landscape of somatic genetic alterations in breast cancers from CHEK2 germline mutation carriers [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P4-04-01.

Abstract P3-06-08: Mucinous and neuroendocrine breast cancers: A spectrum of genetically-related lesions distinct from common forms of estrogen receptor-positive breast cancers

Abstract Introduction: Mucinous (MBCs) and neuroendocrine breast carcinomas (NBCs) are special histologic types of estrogen receptor (ER)-positive breast cancers. Together, MBCs and NBCs account for approximately 5% of invasive breast cancers. NBCs are defined by the expression of neuroendocrine markers (i. e., chromogranin and/ or synaptophysin), whereas MBCs consist of nests of cancer cells floating in pools of extracellular mucin and can be either hypocellular/type A (MBC-A) or hypercellular/type B (MBC-B). MBC-B may display neuroendocrine differentiation. We compared the repertoire of somatic genetic alterations in MBCs-A and MBCs-B, and in NBCs, and evaluated whether they differ from ER-positive/ HER2-negative invasive ductal carcinomas of no special type (IDC-NSTs). Materials and methods: Reanalysis of the targeted capture or whole exome sequencing data of 22 MBCs (12 MBCs-A and 10 MBCs-B), and of 15 NBCs was performed. The BAM files were retrieved and somatic mutations were detected with MuTect, indels with Strelka, Varscan2, Scalpel and Lancet, and copy number alterations using FACETS. The mutational repertoire of MBCs-A and MCBs-B was compared to that of NBCs, and of ER-positive/ HER2-negative IDC-NSTs from The Cancer Genome Atlas (TCGA) breast cancer study (n=310). Results: The most frequently mutated genes in MBCs-A were SF3B1 and FRG1B (17%, each); in MBCs-B were GATA3 and KMT2C (30%, each), and in NBCs were FOXA1 and TBX3 (20%, each) and KMT2C (20%). No significant differences were observed in single gene comparisons between MBCs-A, MBCs-B and NBCs (Fisher's exact tests, p&amp;gt;0.05). When compared with ER-positive/HER2-negative IDC-NSTs from TCGA, NBCs harbored a higher frequency of mutations targeting FOXA1 (20% vs 2.9%, Fisher's exact test, p&amp;lt;0.05) and TBX3 (20% vs 2.9%, Fisher's exact test, p&amp;lt;0.05), and a lower frequency of TP53 mutations (0% vs 23.9%; Fisher's exact test, p&amp;lt;0.05). We observed a lower frequency of PIK3CA mutations in both MBCs-A (8.3% vs 40%, Fisher's exact test, p&amp;lt;0.05) and MBCs-B (0% vs 40%; Fisher's exact test, p&amp;lt;0.01) compared to ER-positive/HER2-negative IDC-NSTs. Although not statistically significant, PIK3CA mutations were numerically less frequent in NBCs than in ER+/HER2- IDC-NSTs (13.3% vs 40%; Fisher's exact test, p&amp;gt;0.05). Concurrent 1q gains and 16q losses, the hallmark copy number alterations of ER-positive/HER2-negative breast cancer were found in 47% of the ER-positive/HER2-negative IDC-NSTs from TCGA, but only present in 20% of NBCs and in none of the MBCs-A or MBCs-B. Conclusion: Type A MBCs, type B MBCs and NBCs harbor similar patterns of genetic alterations, including a low frequency of PIK3CA mutations and of concurrent 1q gains/ 16q deletions. Taken together, our data suggest that both MBCs and the vast majority of NBCs constitute a spectrum of histologically and genomically related subtypes, distinct from ER-positive/HER2-negative IDC-NSTs. Citation Format: Pareja F, Da Cruz Paula A, Selenica P, Lee JY, Marchio C, Brown D, Geyer FC, Brogi E, Vincent-Salomon A, Weigelt B, Reis-Filho JS. Mucinous and neuroendocrine breast cancers: A spectrum of genetically-related lesions distinct from common forms of estrogen receptor-positive breast cancers [abstract]. In: Proceedings of the 2018 San Antonio Breast Cancer Symposium; 2018 Dec 4-8; San Antonio, TX. Philadelphia (PA): AACR; Cancer Res 2019;79(4 Suppl):Abstract nr P3-06-08.

ADS-HCSpark: A scalable HaplotypeCaller leveraging adaptive data segmentation to accelerate variant calling on Spark

BackgroundThe advance of next generation sequencing enables higher throughput with lower price, and as the basic of high-throughput sequencing data analysis, variant calling is widely used in disease research, clinical treatment and medicine research. However, current mainstream variant caller tools have a serious problem of computation bottlenecks, resulting in some long tail tasks when performing on large datasets. This prevents high scalability on clusters of multi-node and multi-core, and leads to long runtime and inefficient usage of computing resources. Thus, a high scalable tool which could run in distributed environment will be highly useful to accelerate variant calling on large scale genome data.ResultsIn this paper, we present ADS-HCSpark, a scalable tool for variant calling based on Apache Spark framework. ADS-HCSpark accelerates the process of variant calling by implementing the parallelization of mainstream GATK HaplotypeCaller algorithm on multi-core and multi-node. Aiming at solving the problem of computation skew in HaplotypeCaller, a parallel strategy of adaptive data segmentation is proposed and a variant calling algorithm based on adaptive data segmentation is implemented, which achieves good scalability on both single-node and multi-node. For the requirement that adjacent data blocks should have overlapped boundaries, Hadoop-BAM library is customized to implement partitioning BAM file into overlapped blocks, further improving the accuracy of variant calling.ConclusionsADS-HCSpark is a scalable tool to achieve variant calling based on Apache Spark framework, implementing the parallelization of GATK HaplotypeCaller algorithm. ADS-HCSpark is evaluated on our cluster and in the case of best performance that could be achieved in this experimental platform, ADS-HCSpark is 74% faster than GATK3.8 HaplotypeCaller on single-node experiments, 57% faster than GATK4.0 HaplotypeCallerSpark and 27% faster than SparkGA on multi-node experiments, with better scalability and the accuracy of over 99%. The source code of ADS-HCSpark is publicly available at https://github.com/SCUT-CCNL/ADS-HCSpark.git.

Whole-exome sequencing data of suicide victims who had suffered from major depressive disorder

Suicide is one of the leading causes of mortality worldwide; it causes the death of more than one million patients each year. Suicide is a complex, multifactorial phenotype with environmental and genetic factors contributing to the risk of the forthcoming suicide. These factors first generally lead to mental disorders, such as depression, schizophrenia and bipolar disorder, which then become the direct cause of suicide. Here we present a high quality dataset (including processed BAM and VCF files) gained from the high-throughput whole-exome Illumina sequencing of 23 suicide victims – all of whom had suffered from major depressive disorder - and 21 control patients to a depth of at least 40-fold coverage in both cohorts. We identified ~130,000 variants per sample and altogether 442,270 unique variants in the cohort of 44 samples. To our best knowledge, this is the first whole-exome sequencing dataset from suicide victims. We expect that this dataset provides useful information for genomic studies of suicide and depression, and also for the analysis of the Hungarian population.

High-resolution mapping of reciprocal translocation breakpoints using long-read sequencing

Long-read nanopore sequencing enables direct high-resolution breakpoint mapping on balanced carriers of reciprocal translocation. The mean sequencing depth on the translocated chromosomes to achieve accurate mapping of breakpoints ranged from 2.5-fold to 6.2-fold. To speed up determination of the breakpoints from long-read sequencing data, alignment reads on the translocated chromosomes were extracted before piped into NanoSV. Checking the position of breakpoints on Interactive Genomics Viewer (IGV) was crucial to successful design of breakpoint PCR primers, especially when large deletion was involved at the breakpoints.•Long-read sequencing enables accurate breakpoint mapping with base-pair resolution•Splitting bam files by translocated chromosomes drastically speeded up the breakpoint determination•IGV helps to identify the breakpoint positions and facilitate the design of breakpoint PCR primers

Data Access and Interactive Visualization of Whole Genome Sequence of Sickle Cell Patients within the St. Jude Cloud

With the increase in availability of high depth whole genome sequencing (WGS) data of individuals with sickle cell disease (SCD), easy access to the raw sequencing data remains an issue due to technical and regulatory challenges. A compliant system that can provide facile data access would accelerate scientific discovery of genetic variants associated with clinical phenotypes. Cloud storage and computing provide an ultimate solution to this data access, which we have shown through the St. Jude Cloud (https://stjude.cloud) where over 5000 whole genome sequences for pediatric cancer patients are being shared in collaboration with DNANexus and Microsoft. Here we expand the St. Jude Cloud to sickle cell disease data through the Sickle Genome Project (SGP) Data Portal (https://pecan.stjude.org/permalink/sgp) to allow instantaneous raw data access (following data access committee approval), as well as visualization of genotype calls at individual level in a novel genome. The SGP WGS data was generated from 871 patients from St. Jude Children's Research Hospital (St. Jude) through the Sickle Cell Clinical Research and Intervention Program (SCCRIP, Pediatr Blood Cancer. 2018 May 24: e27228) and from Baylor College of Medicine (BCM). All study participants provided informed consent for genomic study and data sharing on IRB-approved research protocols.The SGP data portal will have multi-tiered access. All users will have access to a general heat map view which shows anonymized patient clinical values (e.g., fetal hemoglobin (HbF), mean corpuscular volume (MCV), hemoglobin concentration (Hb)) and relevant SCD modifying variants (e.g., Beta-globin locus, MYB, BCL11A, HBA). The GenomePaint bowser allows for viewing coding and noncoding variants. Displayed with each variant will be a visual indication of the median fetal hemoglobin values for patients homozygous for the reference allele, heterozygous, or homozygous for the alternative allele. The browser also displays erythroid specific DNA-accessibility and epigenetic marks and indicates variant that may disrupt erythroid specific transcription factor binding sites (GATA1 and BCL11A). For anonymization purposes, within the genome browser and heat map views, clinical values and the patients age will be binned into ranges when displayed as single or low count values. Lastly, the ProteinPaint tool (Zhang and Zhou, Nature Gen, Dec 29, 2015) will enable visualization and filtering of variants with reference to protein domain and amino acid sequence.To access processed data such as BAM and VCF files for downstream analyses, a user will be required to apply for access which will be adjudicated by a data access committee. Verified researchers will be granted access to clinical data in a manner consistent with the protocol specific informed consent documentation and protocol under which the sequencing was performed. This may include coded clinical and demographic data when specified by the research protocol and informed consentThe SGP data set will be one of the first WGS datasets from primarily African American Sickle cell patients to be made available to clinicians and researchers worldwide. In addition, no SCD-centric data portal exists that contains controlled access to data and provides graphical tools for visual analysis. The combination of the visual tools and ability to download tools provides the scientific community an invaluable resource for studying sickle cell disease. DisclosuresEstepp:Daiichi Sankyo: Consultancy; NHLBI: Research Funding; Global Blood Therapeutics: Consultancy, Research Funding; ASH Scholar: Research Funding. Hankins:Global Blood Therapeutics: Research Funding; bluebird bio: Consultancy; NCQA: Consultancy; Novartis: Research Funding.

Comprehensive mitochondrial DNA analysis and IVF outcome.

STUDY QUESTIONDo mitochondrial DNA (mtDNA) copy number and heteroplasmy in human embryos affect the ongoing pregnancy rate?SUMMARY ANSWEROur study suggests that mtDNA copy number above a specific threshold is associated with the ongoing pregnancy rate.WHAT IS KNOWN ALREADYMitochondria play a vital role in cell function. Recently, there has been increasing research on mtDNA as a biomarker of embryo implantation. Although reports showed that high levels of mtDNA in the blastocyst are associated with low implantation potential, other publications were unable to confirm this. Confounding factors may influence the mtDNA copy number in euploid embryos. On the other hand it has been speculated that both mtDNA heteroplasmy and copy number contribute to mitochondrial function. Next generation sequencing (NGS) allows us to study in depth mtDNA heteroplasmy and copy number simultaneously.STUDY DESIGN, SIZE, DURATIONA prospective non-selection study was performed. We included 159 blastocyst biopsies from 142 couples who attended our clinic for preimplantation genetic testing for aneuploidies (PGT-A), from January 2017 to December 2017. All embryos were biopsied on Day 5 or Day 6. The aneuploid testing was performed by NGS. All blastocysts were diagnosed as euploid non-mosaic and were transferred. The mtDNA analysis was performed once the embryo diagnosis was known.PARTICIPANTS/MATERIALS, SETTING, METHODSSequencing reads mapping to the mtDNA genome were extracted from indexed bam files to identify copy number and heteroplasmy. The relative measure of mtDNA copy number was calculated by dividing the mtDNA reads by the nuclear DNA value to normalize for technical variants and the number of cells collected at the biopsy. All the results were subjected to a mathematical correction factor according to the embryo genome. Heteroplasmy was assigned by MitoSeek.MAIN RESULTS AND THE ROLE OF CHANCEThe mean average copy number and SD of mtDNA per genome was 0.0016 ± 0.0012. Regarding heteroplasmy, 40 embryos were heteroplasmy carriers (26.32%). MtDNA variants were detected in coding and non-coding regions and the highest number of variants in an embryo was eight. With respect to IVF outcome for mtDNA copy number analysis, we set a threshold of 0.003 for the following analysis. The vast majority of the embryos were below the threshold (142/159, 89.31%) and 17 embryos were classified as having higher mtDNA levels. We showed a reduction in ongoing pregnancy rate associated with elevated mtDNA copy number (42.96% versus 17.65%, P < 0.05). This result was independent of maternal age and day of the biopsy: these factors were included as confounding factors because mtDNA copy number was negatively correlated with female age (25 –30 y: 0.0017 ± 0.0011, 30 –35 y: 0.0012 ± 0.0007, 35 –40 y: 0.0016 ± 0.0009, over 40 y: 0.0024 + 0.0017, P < 0.05). Embryos biopsied on Day 5 were more likely to have higher quantities of mtDNA compared with those biopsied on Day 6 (0.0017 versus 0.0009, P < 0.001). According to IVF outcome and heteroplasmy, a lower ongoing pregnancy rate was reported for embryos that carried more than two variants. However, this did not reach statistical significance when we compared embryos with a number of variants lower or higher than two (39.15 versus 20.0, P = 0.188). Finally, a clear positive association between the mtDNA variants and copy number was reported when we compare embryos with or without heteroplasmy (0.0013 ± 0.0009 versus 0.0025 ± 0.0014, P < 0.001) and among different numbers of variants (0:0.0013 ± 0.0009, 1–2:0.0023 ± 0.0012, >2:0.0043 ± 0.0014, P < 0.05).LIMITATIONS, REASONS FOR CAUTIONA limitation may be the size of the sample and the high-throughput sequencing technology that might not have detected heteroplasmy levels below 2% which requires high sequence depth A clinical randomized trial comparing the clinical outcome after the transfer of embryos selected according to mtDNA levels or only by morphological evaluation will be necessary. More research into the impact of mtDNA heteroplasmy and copy number on IVF outcome is needed.WIDER IMPLICATIONS OF THE FINDINGSOur results demonstrate that embryos with elevated mtDNA copy number have a lower chance of producing an ongoing pregnancy. MtDNA copy number is higher in older women and is dependent upon the number of cell divisions that preceded biopsy. Moreover, our data suggest that mitochondrial activity could be a balance between functional capacity and relative mtDNA copy number.STUDY FUNDING/COMPETING INTEREST(S)There are no conflicts of interest or sources of funding to declare.Trial registration numberNot applicable.