Issues with adapter trimming when there are no adapters?

Question

[DareDevil]

When I run fastqc on my sample, it did not identify any adaptor sequence in the sample but there were some low quality reads in the sample. Then I used trim galore to remove the low quality reads using following command:

trim_galore --illumina --gzip --fastqc -o trimmed -j 4 --paired $Read1 $Read2

but in this when I checked the log file I was able to see some adaptors detected in the file. How is this possible? Are there any way to trim only low quality reads?

=== Summary ===

Total reads processed:             115,217,093
Reads with adapters:                40,928,868 (35.5%)
Reads written (passing filters):   115,217,093 (100.0%)

Total basepairs processed: 11,521,709,300 bp
Quality-trimmed:             199,850,659 bp (1.7%)
Total written (filtered):  11,118,622,037 bp (96.5%)

Answer 1

[GenoMax]

Are there any way to trim only low quality reads?

FastQC does not look at every read when it does QC. It sub-samples data. This is fine for QC purposes and done in view of time/memory considerations. Where as a program like trim_galore is doing an operation it looks at every read in your dataset.

If you want to leave reads with adaptors in (which BTW are likely going to carry a lower quality, so you are going to remove some of those) and trim only based on Q scores then you could try bbduk.sh from BBTools. trim_galore likely could be used in a similar mode. You may want to check into command line options.