How should scMLPSeq quality control metrics be interpreted?

Answer:
The scMLP-Seq QC report generated by SeekSoulTools evaluates sequencing performance, probe hybridization and ligation quality, barcode validity, and single-cell expression complexity. Compared with unbiased transcriptome workflows, scMLP-Seq uses a targeted probe-based capture strategy and is therefore particularly sensitive to probe specificity, ligation efficiency, and FFPE sample quality.

The report is organized into three major categories:

• Sequencing QC — evaluates raw sequencing quality and barcode validity. 
• Cell Metrics — evaluates effective cell recovery, sequencing depth, and transcript complexity. 
• Probe Mapping QC — evaluates probe hybridization specificity, ligation accuracy, and usability of reads for downstream quantification. 

For FFPE samples, some transcriptomic metrics (especially median genes per cell and total genes detected) are expected to be lower than fresh-sample scRNA-seq due to RNA fragmentation and targeted capture design. QC interpretation should therefore focus on overall consistency across metrics rather than any single parameter alone.

1. Sequencing QC Metrics

2. Cell-Level QC Metrics

3. Probe Mapping and Ligation QC Metrics

These metrics specifically evaluate probe hybridization accuracy, ligation specificity, and the proportion of reads suitable for downstream quantification.

Recommended QC Interpretation Strategy

scMLPSeq QC evaluation should not rely on a single metric alone. Instead, interpretation is recommended in the following order:

1. Confirm sequencing and barcode quality
   Review Valid Barcode and sequencing saturation first. 
2. Evaluate cell recovery and transcript complexity
   Assess Fraction Reads in Cells, Estimated Number of Cells, and Median Genes per Cell together. 
3. Assess probe hybridization and ligation specificity
   Focus on confidently mapped probe metrics and split/half-mapped read proportions. 

Interpret FFPE samples with appropriate expectations
FFPE-derived samples commonly show reduced transcript complexity compared with fresh samples. Slightly reduced median genes per cell may still be acceptable if probe mapping specificity and cell-associated read fractions remain strong.