On-target editing efficiency

Table 4 displays the percent of observations (efficiency or specificity) that were associated with all On-target loci for All alignments. Further the efficiencies for Pileups, Paired, and Matched criteria are displayed in the following columns. These different criteria are used as the denominator to dictcate the amount of observed nuclease-specific editing. This is an estimate though, as On-target editing does have the potential to saturate the dynamic range of the abundance calculation. Therefore, these percentages should be considered lower bounds for editing efficiency and specificity.

Specimen information

Using the criteria discussed previously based on characterizing features of nuclease targeted sites, off-target sites can be selected from the data in an unbiased manner. Table 6 shows a summary of the unique off-target locations (loci) observed in the data. For All alignments, the loci are based on overlapping alignments (pileup clustering) without a minimum number of fragments required to be classified as a pileup cluster. Pileup loci are similarly based on overlapping alignments, but require at least 3 alignments to form a cluster. Flanking Paired loci require at least two unique alignments with opposite orientation (strands) within 200 bp upstream of each other. Target Matched loci require a match in the upstream sequence to a treated target (within 6 mismatches out of the 18, 20 nts and 1 PAM mismatch).

Off-target enrichment in cancer-associated genes

Flanking Paired loci and Target Matched loci are tested for enrichment against specific gene lists in Table 7. The cancer-associated and special gene lists included in this analysis were: http://bushmanlab.org/assets/doc/allOnco_Feb2017.tsv and http://bushmanlab.org/assets/doc/humanLymph.tsv. Enrichment was tested by Fisher’s Exact and p-values were adjusted for multiple comparisons using a Benjamani-Hochberg correction. Omitted specimens or conditions had insufficient data for this analysis (Total Gene Count = 0) or did not have enough data to support a powerful analysis (Estimated Power greater than 80%).

Genomic distribution of incorporation sites

The figure(s) below display the genomic distribution of identified incorporation sites. The inner most ring plots all alignments identified within the associated data, while subsequent rings plot the alignments associated with Pileups, Flanking Pairs, and Target Matched groups. The height of the bar within its associated ring is correlated to the number of incorporations identified within the 10 Mb window (logarithm base 10 of incorporation site abundances).

Figure 1. Genomic distribution of incorporation sites by bioinformatic characteristics.

Off-target sequence comparison

Off-target sites can be identified by sequence similarity within 100 bp upstream of incorporation sites. The sequences of the top target matched sites are displayed below in Figure 3 for each condition and target along with the number of mismatches to the targeting sequence (Mismatch), an indication if the site is associated with an On- or Off-target location (Target), the total number of unique alignments associated with the site (Abund.), the maximum edit site likelyhood (MESL), and an identifier denoted by the nearest gene (Gene_ID). MESL is a score for the percentage likelyhood the off-target site is associated with directed nuclease editing, based solely on the respective On-target incorporation distribution. The gene name within the Gene_ID is the nearest gene to the genomic location. Further, symbols after the gene name indicate: * that the site is within the transcription unit of the gene, ~ the gene appears on the cancer-association list, ! and that the gene appears on the special gene list. For this report, gene lists used were: http://bushmanlab.org/assets/doc/allOnco_Feb2017.tsv and http://bushmanlab.org/assets/doc/humanLymph.tsv. For a complete list of predicted Off-target sites identified by the analysis, please see the Supporting material section, Off-target site tables.

Figure 2. Sequence similarity between off-target sites and targeting sequence(s).

Editing distributions at On-target loci

Figure 3 displays the distribution of dsODN incorporations around On-target site(s). Incorporations in different orientations are shown on the positive (red) and negative (blue) y-axis. The percentage in the bottom right corner of each plot is an estimate of the number of incorporations associated with the on-target site (based on pileups) captured within the allowed window of 100 bps. These data can be used to fine tune the processing analyses, specifically the upstreamDist parameter which modifies the distance upstream of incorporation sites to search for nuclease edited sequences.

Figure 3. Distance distribution of observed incorporation sites from On-target loci.

Off-target site tables

Below is a complete list of predicted Off-target sites identified for each condition and target. The additional column Edit Site (compared to Figure 3) indicates the predicted genomic editing location, and is broken into three parts delimited by “:”. The first part indicates the chromosome, the second the orientation with respect to genomic numbering (sense “+” or anti-sense “-”), and the third is the location of the targeted base pair.