PRJNA494345

This is a bulk human RNA-seq dataset (Illumina HiSeq 2500, ~1.79 billion reads), and it earns a fully measured grade-A verdict: the read-level quality metrics that actually drive reuse were directly measured, not extrapolated (evidence_strength=1), so the rating is solid rather than provisional. The grade is anchored by an excellent pct_q30_bases of 95.2% alongside a mean base quality of 36.5 — meaning the overwhelming majority of bases are called with high confidence, which is the single most important factor for trustworthy spliced alignment and quantification — while negligible adapter content (0.02%) confirms the reads are clean and ready to map without aggressive trimming. The only metric worth a researcher's attention is the 23.22% duplication rate; this did not lower the score and is entirely normal for deep RNA-seq (where highly expressed transcripts legitimately produce identical fragments), but for applications sensitive to PCR artifacts, such as allele-specific or low-input analyses, you may want to inspect whether it reflects optical/PCR duplicates versus genuine expression. With a 65 bp mean read length, alignment and standard differential-expression reuse are well supported, though the relatively short reads modestly limit isoform-level and novel-junction resolution.