IowaWound

Iowa Wound Pain processing/analysis for Campbell et al. 2024 are listed below. Scripts to reproduce the figures and table values reported in Campbell et al. (2024) can be found here, using provided data including a Phyloseq object generated with this script from Qiime2 output. The raw data are available under NCBI bioproject PRJNA1102692.

(1) Make manifest .tsv’s for important already-demultiplexed reads

Input

• Raw demultiplexed data path from Qi’s pipeline in IowaWoundData/MiSeqV1V3_35/demultiplexed, IowaWoundData/MiSeqV1V3_32/demultiplexed

Outputs

• GH:IowaWound/mappings/Manifest35.tsv
• GH:IowaWound/mappings/Manifest35.tsv

Scripts

• make_manifest_32.sh
• make_manifest_35.sh

(2) Import data into Qiime2 and summarize input

Input

• Raw demultiplexed data path from Qi’s pipeline in IowaWoundData/MiSeqV1V3_35/demultiplexed, IowaWoundData/MiSeqV1V3_32/demultiplexed
• GH:IowaWound/mappings/Manifest35.tsv, GH:IowaWound/mappings/Manifest32.tsv

Outputs

• IowaWoundData/MiSeqV1V3_35/paired-end-demux35.qza
• IowaWoundData/MiSeqV1V3_32/paired-end-demux32.qza

Scripts

• import_data.sh
• summarize_input.sh

(3) FastQC to get quality metrics

Input

• Raw demultiplexed data path from Qi’s pipeline in IowaWoundData/MiSeqV1V3_35/demultiplexed, IowaWoundData/MiSeqV1V3_32/demultiplexed

Outputs

• Full MultiQC output to IowaWoundData/MultiQC_PostDemuxed

Scripts

• fastqc.sh

(4) Selection and empirical tests of truncation lengths for denoising step

The basic principle of truncation length selection was to select a length that met a high enough median PHRED quality cutoff without cutting reads so short that pairs would fail to overlap and would be filtered out in the denoising step of Qiime2. As described in my 2021-11-2 notebook entry, I used ChooseDadaTruncations.R to identify the read lengths such that each read would meet a Phred quality cutoff of 20, 23, and 25. I then tested the parameters associated with each of these cutoff scores for each run using TestDenoise.sh , and plotted the filtered nonchimeric read yield for each set of parameters using FilteringDenoisingStats.R.

Input

• IowaWoundData/MultiQC_PostDemuxed/mqc_fastqc_per_base_sequence_quality_plot_1.txt
• IowaWoundData/MiSeqV1V3_35/paired-end-demux35.qza
• IowaWoundData/MiSeqV1V3_32/paired-end-demux32.qza
• Range of truncation length parameters for MiSeqV1V3_32 chosen in ChooseDadaTruncations.R
• Range of truncation length parameters parameters for MiSeqV1V3_35 chosen in ChooseDadaTruncations.R:

Scripts

• ChooseDadaTruncations.R
• TestDenoise.sh
• FilteringDenoisingStats.R

Results

To maximize the filtered nonchimeric paired reads yielded from the denoising step, and in particular that of MiSeqV1V3_35 (which had much lower read depth than MiSeqV1V3_32), I selected 282 and 247 as truncation lengths for the forward and reverse reads, respectively.

(5) Denoising

Input

• IowaWoundData/MiSeqV1V3_35/paired-end-demux35.qza
• IowaWoundData/MiSeqV1V3_32/paired-end-demux32.qza
• Truncation lengths selected in step 4

Scripts

• denoise.sh

Output

• IowaWoundData/MiSeqV1V3_32/denoising-stats32.qza
• IowaWoundData/MiSeqV1V3_35/denoising-stats35.qza

(6) Visualizations of denoising output

Input

• IowaWoundData/MiSeqV1V3_32/rep-seqs32.qza
• IowaWoundData/MiSeqV1V3_32/table32.qza
• IowaWoundData/MiSeqV1V3_32/denoising-stats32.qza
• IowaWoundData/MiSeqV1V3_35/rep-seqs35.qza
• IowaWoundData/MiSeqV1V3_35/table35.qza
• IowaWoundData/MiSeqV1V3_35/denoising-stats35.qza

Output

• IowaWoundData/MiSeqV1V3_32/rep-seqs32.qzv
• IowaWoundData/MiSeqV1V3_32/table32.qzv
• IowaWoundData/MiSeqV1V3_32/denoising-stats32.qzv
• IowaWoundData/MiSeqV1V3_35/rep-seqs35.qzv
• IowaWoundData/MiSeqV1V3_35/table35.qzv
• IowaWoundData/MiSeqV1V3_35/denoising-stats35.qzv

Scripts

summarize_denoise.sh

(7) Filter and merge samples from the two runs

Input

• table32.qza
  • samples-to-keep-32.tsv
  • samples-to-keep-35.tsv
  • IowaWoundData/MiSeqV1V3_32/table32.qzv
  • IowaWoundData/MiSeqV1V3_35/table35.qzv
  • IowaWoundData/MiSeqV1V3_32/rep-seqs32.qzv
  • IowaWoundData/MiSeqV1V3_35/rep-seqs35.qzv

Output

• IowaWoundData/Qiime2Data/rep-seqs-merged.qza
• IowaWoundData/Qiime2Data/mergedtable.qza

Scripts

filter_merge.sh

(8) Make phylogeny

Input

• IowaWoundData/Qiime2Data/rep-seqs-merged.qza
• IowaWoundData/Qiime2Data/mergedtable.qza

Outputs

• masked-aligned-rep-seqs-merged.qza (alignment)
• unrooted-tree-merged.qza (unrooted tree)
• rooted-tree-merged.qza (rooted tree)

Scripts

• make_phylogeny.sh

(9) Export data from Qiime2 to R-readable data

export_data.sh https://github.com/Grice-Lab/IowaWound/blob/master/scripts/export_data.sh

Outputs

Qiime2Data/RInput/taxonomy.tsv (full taxonomy table with Kingdom-->species) Qiime2Data/RInput/table-with-taxonomy.biom (.biom format OTU/ASV table with Taxa IDs) Qiime2Data/RInput/table-with-taxonomy.tsv (OTU/ASV table with taxa IDs in .tsv format, easy for sweet R script to read) Qiime2Data/RInput/tree.nwk (unrooted tree in .newick form for input into Phyloseq, unifrac distances)