The Microbiome Of Cancer As Seen Through The Lens Of Public Genome Sequence Data

Proceedings: AACR 106th Annual Meeting 2015; April 18-22, 2015; Philadelphia, PAThere are 10 times more bacterial cells in the human body than human cells and various bacteria are known to influence carcinogenesis. Therefore we sought to investigate if publicly available whole genome and whole transcriptome sequencing data generated by large public cancer genome efforts, like The Cancer Genome Atlas (TCGA), could be used to examine the microbial composition of tumors. The Burrows-Wheeler Aligner (BWA) was used to align a selection of Illumina paired-end sequencing data from TCGA data to the human reference genome and all complete bacterial genomes in the RefSeq database. Reads were filtered for low complexity and duplicates and then assigned an Operational Taxonomic Unit (OTU). Through careful consideration of all of the bacterial taxa present in the cancer types investigated, we found an increased abundance of Acinetobacter spp. in acute myeloid leukemia (AML) samples, Mycobacterium tuberculosis complex in ovarian serous cystadenocarcinomas (OV), and Pseudomonas spp. in stomach adenocarcinoma (STAD) samples. Further investigation prompted us to conclude that Mycobacterium tuberculosis complex read pairs were found in the OV samples due to sequencing center batch effects. The same analysis of the AML and STAD samples suggests that Acinetobacter spp. are truly present in AML samples and that Pseudomonas spp. are actually present in the STAD samples investigated. Upon completion of this analysis, we determined that it is possible to identify bacteria in human tumor samples. However, care must be taken to assess bacteria that arrived in the sample via various forms of contamination. It is also important to note that the sequencing library method heavily biases the bacterial taxa identified. One sample sequenced with RNA-Sequencing, whole exome sequencing (WXS), and whole genome sequencing (WGS) had extensively different bacteria reflected in the RNA-Seq data compared to that of WGS and WXS. In the future, a comprehensive approach should be used when evaluating the microbiome of tumor samples via sequencing data. The methods used in this study are valid and useful for identifying the presence of bacteria in human samples, but may not be an exhaustive picture. More weight should be given to this approach in the future when bacterial associations with diseases are suspected.Citation Format: Kelly M. Robinson, Karsten B. Sieber, Kathleen E. Anderson, David R. Riley, Julie C. Dunning Hotopp. The microbiome of cancer as seen through the lens of public genome sequence data. [abstract]. In: Proceedings of the 106th Annual Meeting of the American Association for Cancer Research; 2015 Apr 18-22; Philadelphia, PA. Philadelphia (PA): AACR; Cancer Res 2015;75(15 Suppl):Abstract nr 1534. doi:10.1158/1538-7445.AM2015-1534