Final candidates header explanation

[ksahlin]

The header is on the following format:

> acc + "_” + support + "_" + p_value + "_" + N_t + "_" + delta_size

Here,

• “acc" is the consensus accession.
• “support" is the number of reads in the final consensus cluster that support all the variants that are unique to this cluster. Note that this is not the total number of reads assigned to the cluster as some reads may not support all variants if multiple (see below example)
• "p-value" is the least significant p-value of the statistical tests to other candidates.
• “N_t" is the number of reads used in the statistical test that proveds the p-value. This is usually the number of reads in the cluster itself plus the number of reads in the cluster it is tested against for similarity.
• “delta_size" is the number of positions different to the closest other consensus (edit distance). This is the combination of SNPs and indels. If an indel is of length 3 if will contribute 3 towards delta_size.

Example

If there are two clusters c1 and c2 for 10 and 5 reads respectively, and their total distance is 3 (e.g., one SNP and one indel of length 2). Lets say you are testing whether the SNP and deletion that c2 have with respect to c1 is statistically supported, and the hypothesis test comes back significant (i.e., it is statistically supported), then you will have in c2’s header:

• N_t = 15 (the reads used to check if the clusters should be merged or if they are statistically significant)
• delta_size=3
• p-value: something lower that the threshold for not merging.
• “support” would be however many reads supported both the SNP and the deletion in c2 in the statistical test. This is somewhere between 1 and 5 (although typically towards the cluster size, e.g. 3, 4, or 5).