Inquiry on calculating avg_distance_to_effective_5p_end

[lykstudio]

I am interested in deriving the avg_distance_to_effective_5p_end value utilizing the data from the avg_distance_to_effective_3p_end column.

Is there any plan to provide the avg_distance_to_effective_5p_end metric in the future? If not, could you elucidate the methodology employed in calculating the avg_distance_to_effective_3p_end?

[apldx]

Hello, we don't have a plan for the 5p end (at some point we'd like to add some additional stats, but would probably need to add an option for legacy output etc; it's a good idea!), but I can describe the method for 3p:

First the sum of the proportions of the distances from the 3p end of the query position along the read is calculated and summed over all reads as

sum_3p_distance += (float) std::abs(base->qpos - three_prime_index) / (float) base->b->core.l_qseq;

(three_prime_index is taken from the BAM file via HTSLIB bam_aux_get(base->b, "Zm");)

and then an average is calculated as

stat.sum_3p_distance / stat.read_count;

The relevant lines are here

bam-readcount/src/lib/bamrc/BasicStat.cpp

Line 66 in c7c76e6

sum_3p_distance += (float) std::abs(base->qpos - three_prime_index) / (float) base->b->core.l_qseq;

and here

bam-readcount/src/lib/bamrc/BasicStat.cpp

Line 140 in c7c76e6

s << (float) stat.sum_3p_distance / stat.read_count;