Miseq reporter

Primers and sequence adapters were removed with the Illumina MiSeq Reporter (version 2.5). The sequences were further processed using scripts implemented through the workflow package Quantitative Insights into Microbial Ecology (QIIME) version 1.9.0 (Caporaso et al., 2010 (link)). Individual sequence reads were joined using FASTQ-join (ea-utils, version 1.1.2-537; Aronesty, 2013 (link)), with a maximum number of 3 mismatches and minimum overlap of 6. Reads were demultiplexed and filtered with the minimum acceptable Phred score of 21. Reads were checked for alignment to the human genome (assembly GRCh38) with BLAST (version 2.2.22). Operational taxonomic units (OTUs) were identified with a closed-reference approach against the Silva v119 reference database (Quast et al., 2013 (link)) using the uSearch (version 5.2.236; Edgar, 2010 (link)) algorithm. Chimeric sequences were removed with the blast_fragments approach implemented in identify_chimeric_seqs.py. Taxonomy was assigned to individual OTUs using the RDP Classifier (version 2.2; Wang et al., 2007 (link)) with a minimum confidence of 0.80. The resulting OTU table was imported into R for filtering and statistical analysis. Our workflow is provided in the Supplementary Materials (Supplementary Figure 1). The 16S rRNA gene sequences are available for download from the Short Read Archive (SRA) under accession number SUB1600610.

Sequencing reads were demultiplexed using the MiSeq Reporter (Illumina). Demultiplexed files were subsequently analyzed for base editing activity using a custom workflow combining the SeqKit^{57 (link)} and CRISPResso2 (ref. ^{58 (link)}) packages. See Supplementary Note 6 for additional details. Post-CRISPResso2 analyzed nucleotide frequencies are listed in Supplementary Table 1.

Demultiplexing of the Sequencing reads was done with the MiSeq Reporter (Illumina). Sequencing reads were aligned to the genome using the bowtie2 algorithm and visualized using the Integrative genome viewer. CRISPResso2 was run in with the following settings: CRISPRessoBatch—batch_settings batch.batch—amplicon_seq -p 4 --base_edit -g -wc -10 -w 20. Corrected reads with the base edited therapeutic SNP were calculated by selecting only reads with the intended edit but no indels in the quantification window. Percentages of corrected read and uncorrected reads were plotted using GraphPad Prism 8.3.1 (GraphPad Software, Inc., San Diego, CA).

Primers and sequence adapters were removed with the Illumina MiSeq Reporter (version 2.5). The sequences were further processed using scripts implemented in the R statistical computing environment with the DADA2 (version 1.10.1) package (27 (link)) (scripts are available on GitHub at https://github.com/lakarstens/ControllingContaminants16S). Briefly, sequences were quality filtered and trimmed (forward reads at 230 nucleotides [nt] and reverse reads to 210 nt) prior to inferring amplicon sequence variants (ASVs) with the DADA2 algorithm. ASVs, which group similar sequences together according to a model that considers sequence abundance and sequencing error, were chosen over traditional operational taxonomic units (OTUs) since they have a finer resolution (28 (link)– (link)30 (link)). Chimeric sequences were removed with the approach implemented in the DADA2 package. Taxonomy was assigned for each ASV to the genus level using the RDP Naive Bayesian Classifier (31 (link)) implemented in DADA2 with the SILVA database (version 132). The R package phyloseq (version 1.26.1) (32 (link)) was used for storing the ASV table, taxonomy, and associated sample data and for calculating alpha-diversity measures. Expected values for alpha-diversity measures were calculated on the subset of the mock microbial community dilution samples that only contained expected sequences.

Reads were demultiplexed and adaptors were removed using Illumina Miseq Reporter, according to the manufacture recommendations. Raw sequence reads for all samples were quality filtered using the pair-end mode of Trimmomatic v0.36 [41 (link)]. This software was used to remove low quality bases from the beginning and end of sequence reads pairs (trimming). Also, a sliding window of 8 bases from left to right was performed. Sequence reads were cut whenever the average quality into the window fell below the threshold (<15, Phred score) and the right side of the read sequence was deleted. Sequences with a minimum read length of 150 nt, were retained. Then, the retained paired-reads were merged into a consensus sequence with its associated corrected base quality scores and chimeras were removed using LeeHom software [42 (link)] with default parameters.