2.3.3. ITS-Based BLAST Performance

We used a subset of 758 terminals with (near) complete ITS sequences as reference (File S5, "BLAST reference subset"), similar to the aforementioned *Cora* subset but also including *Acantholichen* and *Corella*, to assess BLAST performance of four datasets: (1) the ITS1 region only (298 bases, individual length (173–)200–220 bp) of the *Cora* species in the BLAST reference subset (716 terminals; File S6); (2) the ITS region of all original terminals with less than 10% gaps (including Illumina ASVs, 1217 terminals; File S7); (3) the ITS2 region only (344 bases, individual length 231–263 bp) of this BLAST reference subset (File S8); and (4) a short, subterminal (49 bp from the end of the ITS4 primer) piece of the ITS2 region (85 bases, individual length (43–)50–64 bp) of this BLAST reference subset (File S9). The analysis was performed using the local BLAST function in BioEdit 7 [69,72] with the following settings: E value = 1.0, maximum number of hits to report = 50, gap opening/extending penalty = 0, mismatch penalty = −3, match reward = 1 (see also [73,74]).

BLAST results for each of the scenarios were evaluated as follows. First, we computed the variation of total scores for hits (same species-level clade), self-hits (same terminal), and misses (different species-level clade) for each scenario. Then, we computed the minimum score for hits versus the maximum score for misses for each query, determining to what level a "BLAST gap" existed for each scenario. Finally, we computed the ratio of minimum hit score versus maximum miss score to calculate the proportion of three ratios per query: (1) >1.25, minimum hit score at least 25% higher than maximum miss score; (2) >1.10, minimum hit score at least 10% higher than maximum miss score; and (3) >1.00, minimum hit score higher than maximum miss score.
