Hi, when I set a top percent filter, the reads are not being classified using it. Any idea what’s going on? If im understanding it correctly, then in the following pictures, the lowest possible bit score of a match being used should be 397.8, if the highest bit score of the read is 442 (see attached image), though it is still using those matches to classify the read.
When using the long read LCA, the top-percent criterion is used locally, that is, it is applied only to overlapping segments of alignments only. The shading in the inspector viewer does not currently show that, that is why all appear in black. Use the naive LCA for global filtering of alignments.