HUMAnN3: single-end reads equal to 50 bp, resulting in zero species

Microbial community profiling MetaPhlAn
1

I’m encountering complete classification failure using MetaPhlAn4/HUMAnN3 with single-end 50bp metagenomic data. Despite proper database configuration and parameter tuning, all reads remain unclassified. Has anyone resolved similar issues?
For example:
@S250118066L1C001R00100001100
CTATGATTGGAAGCTTCTTGAGACCTCCCCAGAAGCAGATGCAAGTCGGA
+
FDECFADDFFCAFF@BFDDFCFCFF@FFFFCF@EFFEFA@FFCEFAFFFC