The purpose of this paper was to compare next generation sequencing (NGS) results obtained from archival formalin-fixed paraffin-embedded (FFPE) tumor specimens using two different library preparation TruSeq kits by two different commercial laboratories that performed RNA extraction, library construction, and NGS. The authors also examined potential effects associated with specimen type (core needle biopsy versus resection), tumor type, and block storage duration on RNA yield and integrity.
Comparison of Two Illumina Whole Transcriptome RNA Sequencing Library Preparation Methods Using Human Cancer FFPE Specimens.
Wang D, Rolfe PA, Foernzler D, O'Rourke D, Zhao S, Scheuenpflug J, Feng Z
Purpose
Conclusion
The percentage of RNA fragments longer than 200 bp (DV200) was correlated with the duration of FFPE block storage when FFPE specimens underwent RNA extraction, library construction, and NGS analysis by one of the vendors used, while an association between library yield and FFPE block age was observed with the other vendor. RNA yields were comparable between surgical resection and core needle biopsy specimens and between the cancer types evaluated.
Alignment rates, and the percentage of exonic and intragenic reads were higher for libraries constructed using the TruSeq RNA Access kit than the TruSeq Stranded Total RNA kit. Correlations among sequencing results generated by two different vendors were strong for both library preparation methods. Between-vendor correlations were stronger when library concentrations were used for comparisons than when RNA concentrations were used While sequencing results from different vendors were generally strongly correlated between library preparation methods, a bias was observed for highly expressed genes.
Studies (1)
Purpose
The purpose of this study was to compare NGS results obtained by two different vendors from archival FFPE tumor specimens using two different library preparation TruSeq kits. The authors also examined potential effects associated with specimen type (core needle biopsy versus resection), tumor type, and block storage duration on RNA yield and integrity. FFPE blocks from 25 patients with non-small cell lung cancer, colorectal cancer, renal carcinoma, breast cancer, or hepatocellular carcinoma (numbers of each not specified) were obtained from commercial suppliers. Twenty-two of the specimens were from surgical resections and three were from core needle biopsies; no other details of collection were provided. The FFPE blocks were stored for 3-25 years prior to the study; storage temperature and condition were not reported. RNA was extracted from five 5 µm-thick sections using the Qiagen miRNeasy FFPE kit by two commercial laboratories. RNA was quantified using the Qubit RNA HS Assay Kit and integrity was assessed using the Agilent RNA 6000 Nano Kit on a 2100 Bioanalyzer. Sequencing libraries were constructed by both vendors with the TruSeq Stranded Total RNA library preparation kit with Ribo-Zero Gold, or the TruSeq RNA Access library preparation kit. Libraries were sequenced on an Illumina HiSeq 2500 machine using V3 chemistry in paired-end mode with a read length of 2 x 50 bp. One vendor sequenced TruSeq Access libraries to a minimum of 100 million reads and Total RNA libraries to 50 million reads and the other vendor sequenced both library types to a minimum of 100 million reads per sample. Data were processed using a standard RNASeq pipeline.
Summary of Findings
Of the 25 specimens evaluated, 23 yielded sufficient RNA for TruSeq Stranded Total RNA library preparation and 21 had sufficient RNA for a second library preparation with the TruSeq RNA access kit by one of the two vendors used for RNA extraction, library construction and sequencing. Of these 21 specimens, libraries from 19 specimens were suitable for sequencing (>2 nm) by both methods by the first vendor. The second vendor started with 22 specimens, all of which produced sufficient libraries for sequencing using the TruSeq Stranded Total RNA library preparation method; 20 of these specimens had enough RNA remaining to prepare libraries using the TruSeq RNA access kit, with 16 specimens yielding libraries sufficient for sequencing. The authors report that the percentage of RNA fragments longer than 200 bp (DV200) was correlated with the storage duration of the FFPE block; however, no correlation coefficients were reported, and the relationship was only statistically significant at one vendor (P=0.015 and P=0.51). Interestingly, the vendor that did not find a correlation between DV200 values and FFPE storage, observed an association between library yield using the TruSeq Access library preparation kit and block age (P=0.02), which was not observed by the vendor who did observe a correlation between block age and DV200 values. RNA yield was comparable between surgical resection and core needle biopsy specimens and between the cancer types evaluated.
Alignment rates were higher for libraries constructed using the TruSeq RNA Access kit than the TruSeq Stranded Total RNA kit for both vendors (95% and 93% versus 83% and 78%, respectively). Compared to TruSeq Stranded Total RNA kit libraries, TruSeq RNA Access kit libraries also produced a higher percentage of exonic reads (81% and 84% versus 18.5% and 41.6%, respectively), intragenic reads (89.1% and 92.1% versus 83.1% and 81.5%, respectively). Spearman’s and Pearson’s correlations in sequencing results between vendors were strong for both the TruSeq Stranded Total kit (ρ =0.87, and r=0.76, respectively) and the TruSeq Access kit (ρ=0.89 and r=0.73, respectively). Between-vendor correlations were more strongly correlated with library concentrations than RNA concentrations (ρ=0.732-0.964 versus ρ=0.264-0.481, respectively). Between-protocol Spearman and Pearson correlations were 0.81 and 0,13, respectively, for one vendor and 0.83 and 0.22, respectively, for the other vendor. Importantly, signals were higher for highly expressed genes when the TruSeq RNA Access kit was used compared to the TruSeq Stranded Total kit.
Biospecimens
TissueBreastTissueColorectalTissueLungTissueKidneyTissueLiverPreservation Types
FormalinDiagnoses
NeoplasticCarcinomaPlatforms
RNANext generation sequencingRNAAutomated electrophoresis/BioanalyzerRNAFluorometryPre-analytical Factors
Classification Pre-Analytical Factor Value(s) Preaquisition Diagnosis/ patient condition Breast cancerColorectal cancerHepatocellular carcinomaNon-small cell lung cancerRenal carcinomaBiospecimen Acquisition Method of tissue acquisition Core needle biopsySurgical resectionStorage Storage duration 3-25 yearsAnalyte Extraction and Purification Analyte isolation method Between site variabilityNext generation sequencing Specific Template modification TruSeq RNA Access library preparation kitTruSeq Stranded Total RNA library preparation kit with Ribo-Zero Gold