Table 4 Summary of the methods used by the different laboratories.
Methods | Lab | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | |
Processing of raw reads | Cutadapt—trimmed primers | sickle—qual trim Q35 | usearch fastx_truncate—trim primers | None | None | None | None | Trimmomatic—qual trim < 20 Phred score | Trimmomatic—qual trim < 3 Phred score | BIOiPLUG MTB (entirely black box) | None | Primers trimmed. Reads trimmed at Phred quality score < 3. Removed if length < 30 bp | None |
Paired end read join | Mothur make.contigs | QIIME1 fastq-join | QIIME1 fastq-join | Mothur make.contigs | Mothur make.contigs | Mothur make.contigs | FLASH | QIIME1 fastq-join | Seqprep | BIOiPLUG MTB | USEARCH fastq_mergepairs | QIIME1 fastq-join | Mothur make.contigs |
Removed reads with "N" | Yes | Yes | Yes | Yes | Yes (> 4) | Yes | No | Yes | Yes (> 10%) | BIOiPLUG MTB | No | Yes | Yes |
Removed reads by length | Yes (removed outlier small and long reads) | No | No | Yes (removed outlier small and long reads) | Yes (removed outlier long reads) | No | No | No | Removed reads < 100 bp | BIOiPLUG MTB | Removed reads < 250 bp | No | Yes, removed outlier long and short reads |
Other pre cluster processing | Removed reads with a homopolymer run longer than 8 | Removed sequences that didn't have primer sequence present, truncated reads at Q score < 19 | Truncated reads at Q score < 19 | Removed reads with a homopolymer run longer than 6, removed sequences that didn't have primer sequence present | Removed reads with a homopolymer run longer than 9 | Removed reads with a homopolymer run longer than 8 | None | None | None | BIOiPLUG MTB | Removed sequences with greater than 0.5 expected errors, truncated reads at Q15 | None | Removed reads with a homopolymer run longer than 6 |
pre OTU read alignment | Mothur against SILVA database | None | None | None | Mothur against SILVA database | Mothur against SILVA database | None | None | Infernal against RFAM, followed by MAPseq against SILVA | BIOiPLUG MTB | None | None | Mothur against SILVA database |
Pre cluster | Mothur ~ 99% similarity | None | None | None | Mothur ~ 99% similarity | Mothur ~ 99% similarity | None | None | None | BIOiPLUG MTB | None | None | Mothur ~ 99% similarity |
Chimera detection | VSEARCH | VSEARCH | USEARCH61 | USEARCH61 | UCHIME | UCHIME | None | USEARCH61 | None | BIOiPLUG MTB | None | None | UCHIME |
OTU assignment and database | Mothur de novo | QIIME1 UCLUST against SILVA | QIIME1 USEARCH de novo | QIIME1 UCLUST against SILVA | Mothur de novo | Mothur de novo | BLASTN against RefSeq RNA database | QIIME1 USEARCH de novo | MAPseq against SILVA | BIOiPLUG MTB | USEARCH de novo OTU clustering | UCLUST OTU clustering against Greengenes. USEARCH de novo clustering against failures | Mothur de novo |
Taxonomic classifcation and database | Mothur against RDP database | QIIME1 UCLUST against SILVA | QIIME1 RDP against SILVA | QIIME1 PYNAST aginst SILVA | Mothur against custom database | Mothur against SILVA database | BLASTN against RefSeq RNA database | QIIME1 UCLUST against SILVA | MAPseq against SILVA | BIOiPLUG MTB | QIIME1 UCLUST against Greengenes | QIIME1 UCLUST against Greengenes | Mothur against SILVA database |
Post-processing | Sequences removed if less than 90 reads in 50% of samples | OTUs not represented in all 3 repeats were discarded | QIIME1 removed sequences in table present at fraction 0.005 of total number of reads | QIIME1 removed sequences in table present at fraction 0.001 of total number of reads | None | Oligotyping and ARB | None | QIIME1 removed sequences in table present at fraction 0.00005 of total number of reads | None | BIOiPLUG MTB | Took forward only OTUs with species assignment | None | OTUs kept where mean relative abundance within replicates > 0.0001 |
Normalisation of reads | Percentage abundance per sample | Percentage abundance per sample | Percentage abundance per sample | Percentage abundance per sample | Percentage abundance per sample | Percentage abundance per sample | Percentage abundance per sample | CSS | Percentage abundance per sample | BIOiPLUG MTB | Percentage abundance per sample | Percentage abundance per sample | Percentage abundance per sample |
