Low coverage and low identity together usually means the read is bad, not that the analysis is wrong. A clean COI barcode hits GenBank or BOLD at 95-100% over most of its length even when your exact species isn't in there, because congeners still match. So I'd stop adjusting software settings and look at the trace itself.

Two things to check on the chromatogram. The first 20-40 bases and the tail are always unreliable, so if you're BLASTing the raw read with primer sequence still attached, coverage and identity both drop for reasons that have nothing to do with your insect - trim to the clean window and BLAST that. And doubled peaks running through the whole trace mean you amplified more than one template, which happens a lot with whole-body extractions of small insects. That one is a bench problem and no amount of analysis will fix it.

Disclosure, I work on SeqBench: https://seqbench.com/tools/sanger-trace-viewer opens an .ab1 in the browser, shows per-base Phred quality and exports the trimmed read as FASTA. Nothing to install or misconfigure, which sounds like the CLC issue you're hitting. Any ab1 viewer does the same job.

I think the main issue pertaining to CLC is this "I am familiar with CLC software and have tried using it, but during training or repeated analyses, the settings seem to change and I am not getting consistent or correct results. When I BLAST the sequences on NCBI, the query coverage and identity percentages are always very low, so I am not confident about the results. "

The above user brings up some good points:

"Low coverage and low identity together usually means the read is bad, not that the analysis is wrong. A clean COI barcode hits GenBank or BOLD at 95-100% over most of its length even when your exact species isn't in there, because congeners still match. So I'd stop adjusting software settings and look at the trace itself. Two things to check on the chromatogram. The first 20-40 bases and the tail are always unreliable, so if you're BLASTing the raw read with primer sequence still attached, coverage and identity both drop for reasons that have nothing to do with your insect - trim to the clean window and BLAST that. And doubled peaks running through the whole trace mean you amplified more than one template, which happens a lot with whole-body extractions of small insects. That one is a bench problem and no amount of analysis will fix it."

Were the amplicons cloned into a vector before sequencing? This can help ensure that a single template is sequenced.

Read QC will be important. If raw reads are being used as input to BLAST, then the vector is still on the reads and it will need to be trimmed away. This can be done in CLC using the trim sequences tool: https://resources.qiagenbioinformatics.com/manuals/clcgenomicsworkbench/current/index.php?manual=Trim_sequences.html

The "Trim contamination from sequences " option can be used to specify the vector sequences used for cloning. Trim using quality scores will be useful in removing the low quality base calls near the ends of the reads (which is typical for sanger reads). CLC uses the same BLAST algorithm as NCBI. There are options to change the alignment parameters as needed:

It might be worth running the Trim and Map Sanger sequences template workflow:

https://resources.qiagenbioinformatics.com/manuals/clcgenomicsworkbench/current/index.php?manual=Trim_Map_Sanger_Sequences.html

This workflow will remove low quality base calls/vector from the reads and map them to a reference sequence. You can use the mitochondrial COI gene as the reference. Once the alignment finishes, you can inspect the read mapping file to see if your primers are amplifying the right product.

I hope this helps, please reach out to TS-Bioinformatics@qiagen.com with any questions