Quality control recommendations for RNASeq using FFPE samples based on pre-sequencing lab metrics and post-sequencing...

Area of Science:

• Biomarker discovery
• Molecular pathology
• Genomics

Background:

• Formalin-fixed, paraffin-embedded (FFPE) tissues are valuable for biomarker identification due to availability and long-term follow-up.
• RNA extracted from archival FFPE samples often has limited quality, posing challenges for molecular analysis.
• Identifying parameters for successful RNA extraction and RNA sequencing (RNAseq) from FFPE tissues is crucial.

Purpose of the Study:

• To identify critical parameters for successful RNA extraction, library preparation, and RNAseq data generation from FFPE samples.
• To optimize library preparation protocols specifically for FFPE tissues.
• To develop a predictive model for assessing the quality of FFPE samples for downstream bioinformatics analysis.

Main Methods:

• Optimized library preparation protocols for FFPE samples using FFPE and Fresh Frozen tissue pairs.
• Tested optimized protocols on 130 FFPE breast tissue biopsies.
• Collected RNA extraction and preparation metrics, compared them with bioinformatics sequencing data, and built a decision tree model to predict RNAseq quality control (QC) status.

Main Results:

• FFPE samples failing bioinformatics QC had low correlation, few mapped reads (<25 million), or few detectable genes (<11,400).
• QC-failed samples showed significantly lower median RNA concentration (18.9 ng/ul) and pre-capture library Qubit values (2.08 ng/ul) compared to QC-passed samples.
• A decision tree model using RNA concentration and library Qubit values achieved an F-score of 0.848 for predicting QC status.

Conclusions:

• Established bioinformatics quality control recommendations for FFPE breast tissue samples.
• Recommended minimums for successful RNAseq: 25 ng/ul FFPE-extracted RNA for library prep and 1.7 ng/ul pre-capture library output.
• These metrics ensure adequate RNAseq data for downstream bioinformatics analysis.