Data processing and quality control¶
FastQC (ver 0.11.8)
was used to evaluate the overall quality of the downloaded data, followed by the use of
to remove vector sequences and low-quality bases. Sequences shorter than ⅔ of the original read length were removed from subsequent analysis. The remaining sequences were referred as to
clean data and used for subsequent analysis.
Seqtk was used to convert sequences from FASTQ to FASTA.
Two rounds of quality control procedures were applied. First, for the 16S data, samples (runs) with less than 20,000
clean reads were removed from subsequent analysis, and marked as "failed QC (QC status == 0)" in
Then, after taxonomy assignment, samples contain only a single taxon (i.e., a species or genus accounts for more than 99.99 percent of total abundance) will be marked as "failed QC (QC status == 0)".
Taxonomic assignment to the sequenced reads¶
16S amplicon sequences, QIIME2 was used to analyze the obtained
clean data and assign taxonomic classification information to the reads;
ASV (Amplicon Sequence Variant) instead of
OTU (Operational Taxonomic Units) results were used, as the former can provide more precise measurement of sequence variation and be able to easily compare sequences between different studies (tractability and reproducibility) Wikipedia: ASV . Relative abundances were then calculated for each sample, with the totaling abundance values of 100% respectively.
For whole genome (i.e., metagenomic or mNGS) sequences, MetaPhlAn2 was used with default parameters for the taxonomic classification of the sequencing reads.
Identification of taxon co-occurrence¶
Co-occurred taxa can be functionally relavent. In GMrepo, co-occurred taxon pairs were identified for health and each disease separately.
Two taxa (i.e., either two species or genera) will be considered as co-occurred in a disease (e.g., Crohn Disease) associated samples if they meet all the following criteria:
Fisher's exact testwas used to calculate the odds of the two taxa to co-occur in the disease associated samples/runs based on their presence/absence information. A taxon was considered as present in a sample/run if its relative abundance was higher than 0.01%. A p-value and odds ratio (OR) were reported for each pair. The pair with p-value < 0.05 was selected. The
fisher.test()function implemented in R was used.
Peasrson correlationanalysis was applied to the relative abundances of the two taxa in the disease associated samples. The pair with a p value < 0.05 was selected.
Spearman correlationanalysis was applied to the relative abundances of the two taxa in the disease associated samples. The pair with a p value < 0.05 was selected.