...
Current as of Oct 5, 2020
Sampling and sample processing
This is likely gonna differ by project. Just as a reminder to us all, we should include which kits we used to do extractions, that we used an Integra AssistPlus robot, that we ground the hell out of stuff with a tissue lyser, and that we included a bunch of extraction blanks. Of course, we will all also have a bunch of other sample prep stuff to mention here too.
Library preparation
Prior to library preparation, a synthetically designed internal standard (ISD) was added to extracted DNA. This ISD is described in Harrison et al. 2020 and allows for conversion of the relative abundance data obtained from the sequencer into estimates of actual abundances. To account for cross-contamination, ‘coligo’ sequences were also added to each well (Harrison et al. xx). Coligos are synthetically designed DNAs. By adding a unique coligo to each well, it is possible to track incidences of cross-contamination. We included negative controls within our library to account for contamination of PCR reagents. We also performed library preparation on a ZymoBiomics mock community, as a positive control.
...
Libraries were sequenced by Psomagen (Rockville, Maryland, USA) on an Illumina NovaSeq 6000 using 2x250 paired-end sequencing.
Bioinformatics
Sequence data were demultiplexed using a custom perl script (created by C. Alex Buerkle). Unique reads were identified ('dereplicated') using vsearch v.2.9.0 (Edgar 2010, Rognes et al. 2016). Dereplicated reads were clustered using the ‘cluster_unoise’ (Edgar 2016) algorithm and a 99% similarity threshold. We stipulated that a sequence must occur 12 or more times for it to be considered as a potential OTU. This choice was made because of the very large number of reads we obtained, as a way to avoid analyzing variants caused by technical error. Chimeric sequences were removed using 'uchime3_denovo' algorithm (Edgar et al. 2011) and the resulting OTUs used to make an OTU table using the 'usearch_global' algorithm.
OTUs that corresponded with the ISD were identified using ‘usearch_global' with the ISD sequence as the queried database. Similarly, coligo sequences were identified using the ‘search_exact’ algorithm of vsearch with coligo sequences as the database. Identification of coligos required the ‘search_exact’ algorithm because the heuristics of the 'usearch_global’ algorithm caused occasoinal mismatches during testing, because of how short the coligo sequences were. Computing was performed using the Teton Computing Environment at the Advanced Research Computing Center, University of Wyoming, Laramie (https://doi.org/10.15786/M2FY47).
Citations:
Edgar, R. C. (2010). Search and clustering orders of magnitude faster than BLAST. Bioinformatics, 26(19), 2460-2461.
...