Interpreting Omic Data for Microbially Influenced Corrosion: Lessons from a Case Study Involving a North Slope Production SystemMicrobiologically influenced corrosion (MIC) is a significant source of pitting corrosion affecting oil and gas pipelines wells and a variety of surface facilities. Understanding of MIC is greatly enhanced through DNA and protein sequencing technologies. Both shotgun and targeted investigations can provide useful information regarding the presence and type of MIC occurring. However the methods used to generate the data the data quality and the limitations associated with data interpretation should not be underestimated. In this presentation listed topics are highlighted through a case study involving the metagenomics and proteomic analysis of pig envelope debris and seawater samples from various locations within a North Slope production system suspected to be suffering from MIC.From sample receipt through data interpretation each step of the process can impact the final results. While one set of protocols is not necessarily universally used we have developed Standard Operating Procedures (SOPs) to ensure that samples are analyzed in a consistent manner with the appropriate quality controls. Sample extraction and preparation for sequencing is a key step in any environmental sample examination particularly for metagenomics and proteomics. For metagenomic processing purified DNA can be sequenced (via direct sequencing or indirect hybridization methods) with or without prior target gene amplification. In this study sequencing was performed both with and without 16S rDNA gene amplification. Following bioinformatics testing the resulting data showed dramatically different results when comparing the 16S sequence data to the shotgun-based sequence data. It can be speculated that 16S gene sequencing captures broad shifts in community diversity over time but can suffer from amplification bias. Shotgun-based metagenomic approaches provide robust estimates of microbial community composition and diversity without the need to target and amplify a specific gene but more sequence data are needed. Furthermore phylogenetic classification of microorganisms using shotgun sequencing is seldom coupled to 16S-based classification and a few recent studies have identified discrepancies between the different classification methods usually with regard to the level of resolution obtained. In this case study metagenomic contigs 16S gene sequences encoded in metagenomic data and 16S rDNA amplicon sequences were sometimes concurrent especially for taxa that were both abundant in the sample and well-represented in 16S and genome databases while other times they provided a vastly different picture of microbial community composition and dynamics.Further for shotgun-based metagenomic analyses that seek to identify organisms and genes based on sequence alignment to publically-available databases the source and comprehensiveness of the database is critical. In this study we showed that the difference between using a RefSeq (NCBI) downloaded in 2013 versus an updated database (2015) significantly impacted data interpretations. One particular organism Sedimenticola selenatireducens was found to dominate the relative abundance of the samples when the updated database was used while it was not identified when the 2013 database was used. The presence of this organism was confirmed by genome coverage from the raw metagenomic reads as well as proteomic identification. Similar to DNA sequence interrogation (and perhaps more so) we and others have also found that sample preparation quantification and analysis techniques can greatly impact protein sequencing results.To summarize our case study has illuminated several factors that should be considered if omic-based examinations are used for MIC impacted sites. Differences in extraction methods sample preparation sequencing platforms and the complexity of the samples studied may lead to different or biased observations. In the end quality assurance checks should be performed from sample collection through analysis in order to make the most informed decisions from the resulting data.

 Keywords: metagenomics, proteomics, microbially influenced corrosion, sequencing, bacteria