Figure 2

Read alignment optimization. Barplots of the primary (best matching library entry) and secondary alignment score (second-best matching library entry) achieved by each read. By adjusting the matching threshold, the user can include reads with non-perfect matches to increase sensitivity. By adjusting the ambiguity threshold, the user can control to what degree the algorithm should tolerate reads matching multiple library entries. Reads located on the diagonal match multiple library entries equally well (primary score = secondary score) and are discarded by PinAPL-Py’s default setting (As an example, the popular Human GeCKO library contains almost 4,000 ambiguous sequences matching multiple target genes which will, consequently, be discarded by PinAPL-Py). (A) High stringency setting requires perfect matching and allows no (not even low) matching to another library sequence. (B) The default setting requires perfect matching, but also accepts reads if they have a second-best match in the library, as long as the second-best score is lower than the primary score. (C) High sensitivity setting accepts reads with a less than perfect sequence match, e.g. when accounting for less than perfect sequence quality.