Quality check for SRP265240
Verdanta pipeline version 0.0.1.
Run 2025-12-20 21:28.
Read coverage along transcripts
Coverage along normalized transcript length (0-100%). Forward, reverse, and sum traces are shown per sample x transcriptome; use this to spot 5' or 3' bias from degradation, poly(A) selection, or fragmentation choices.
Sources: RSeQC, Wang2012
Expression distributions (TPM)
Per-transcriptome across samples; TPM on a log scale with box/density summaries. Highlights global shifts, dispersion, and outliers in quantification.
Sources: Salmon docs, Patro2017
Sample expression densities
Within each transcriptome, per-sample TPM density curves on a log scale. X - log TPM, Y - density. Transcripts with no reads are counted separately. Look for global shifts, bimodality (batch effects), or heavy tails from low-count inflation.
Sources: Salmon docs, Patro2017
Read mapping summary
Per sample x transcriptome table of alignment outcomes: unique %, multi-mapped %, and unaligned (if available). Cross-check with the multiplicity and mapped-count plots.
Sources: MultiQC, Salmon docs
Mapped reads (counts)
Total mapped read counts per sample by transcriptome. Use to verify depth uniformity and to spot lane/sample dropouts.
Sources: Salmon docs
Sample-to-sample correlation heatmaps.
Sample-to-sample correlation heatmaps. Each tile displays the Pearson correlation between samples, calculated from gene-level TPM values after applying log(1+x). Undefined pairs are shown as hatched cells.
Sources: Salmon docs, Salmon paper
Adapter content (%)
Estimated adapter sequence content by read position. Rising 3' signal indicates incomplete trimming; strong early-cycle signal can imply short inserts.
Sources: FastQC
Duplication levels
Distribution of duplication counts vs % reads. High duplication implies low library complexity or over-amplification; interpret in light of depth and any UMI handling.
Sources: FastQC
Read length distribution
Distribution of read lengths per sample. A single sharp peak at the expected read length indicates uniform reads; broader or truncated modes suggest trimming or differing run settings (e.g., 2x100 vs 2x150).
Sources: FastQC
Per-base N content (%)
Fraction of 'N' calls at each read position. End-biased N indicates low signal at read termini; pervasive N suggests run chemistry or filtering issues.
Sources: FastQC
Per-base nucleotide composition
A/C/G/T proportion by read position. Early-cycle biases often reflect random-hexamer priming; strong cycle-wide deviations can indicate contamination or low-diversity libraries.
Sources: FastQC
Per-base quality (Phred)
Quality score summaries vs read position. Gradual 3' decay is expected; sharp drops flag chemistry issues or the need for stricter trimming.
Sources: FastQC
Per-sequence GC content
GC% distribution across reads; compare to the organism/transcriptome expectation. Shifts, extra peaks, or broadened spread suggest contamination or selection bias.
Sources: FastQC
Per-sequence quality distribution
Distribution of mean Phred scores per read. A left tail marks subpopulations of low-quality reads that may warrant tighter filtering.
Sources: FastQC