Sample type, sample preparation and sample QC

We have extensive experience with various sample types, including human, mouse and rat samples, ranging from fresh frozen tissues and cell cultures to formalin-fixed paraffin embedded tissues (FFPE) and body fluids, also referred to as liquid biopsies (e.g. serum, plasma, urine, etc).

RNA and sequencing library quality is monitored by using microfluidic / capillary gel electrophoresis (Agilent 2100 Bioanalyzer and AATI Fragment Analyzer) and quantified for all sequencing related applications by Qubit fluorometric quantification (Thermo Fisher), followed by data quality control, to ensure accurate and reproducible results.

Sequencing technologies

Small RNA / miRNA sequencing

Small RNA sequencing enables the analysis of small non-coding RNA molecules, typically in the range of 20-40 nucleotides. The small RNA sequencing work-flow focusses on microRNAs (including isomiRs) as the most prominent class of small RNAs, by specifically size selecting the miRNA fraction using Pippin Prep technology (Sage Sciences). If required, Biogazelle can also actively deplete unwanted small RNA molecules from the total RNA, for instance tRNA fragments in RNA extracted from human and mouse serum, or Y-RNAs in RNA from human plasma (Van Goethem et al., 2016).

Sequencing reads are processed using Biogazelle’s proprietary small RNA sequencing analysis pipeline Cobra (see below) and annotated according to the latest miRBase database. In addition to microRNAs, other small RNAs such as piRNAs, tRNA and sn(o)RNAs (fragments) are also quantified.

With Biogazelle scientists being the study coordinators of the microRNA quality control study (miRQC, Mestdagh et al., 2014), Biogazelle has a lot of expertise and comparative performance know-how. For more information on our small RNA sequencing workflow, take a look at our tech notes and related key presentations.

Stranded total RNA sequencing

Total RNA sequencing results in a comprehensive analysis of both coding and non-coding RNAs, regardless of polyadenylation status of the transcript, and includes all long non-coding RNAs (lncRNAs). As the entire transcriptome is considered, the most complete expression profiling is ensured. Human lncRNA annotation is based on the latest version of LNCipedia, offering one of the richest databases of human lncRNAs. Predicted functions of lncRNAs are available in our proprietary LNCarta database or can be determined using dedicated guilt-by-association studies.

As with polyA+ sequencing, a strand-specific protocol is used, which provides important information with regard to strand orientation of the transcribed RNA, especially important when studying sense/antisense overlapping expression.

For more information on our total RNA sequencing workflow, take a look at our related tech notes.

Stranded polyA+ RNA sequencing

PolyA+ RNA sequencing enables expression analysis of mature polyadenylated RNAs, including the majority of all protein coding mRNAs. By enriching polyA+ RNA, a higher sequencing depth of these RNAs is obtained, resulting in higher sensitivity and more accurate quantification, especially of low abundant genes.

In addition to gene expression analysis, transcript expression analysis, gene fusion detection, mutation analysis and allele-specific expression analysis can be conducted.

A strand-specific protocol is used, which provides important information with regard to strand orientation of the transcribed RNA.

RNA capture sequencing

RNA capture sequencing enables the study of gene expression of low quality human samples, such as FFPE and body fluids. Sequence specific probes covering more than 98% of the human exome allow to focus on the coding regions of mRNA from these challenging samples to maximize the sensitivity and sequencing efficiency.

A subtype of the RNA capture sequencing approach is also available, targeting 1385 cancer-related transcripts, further increasing the sequencing efficiency and offering a cost-effective solution in the oncology field to study both expression levels and structural RNA changes (alternative splicing, mutations, and fusion genes).

For more information on our RNA capture sequencing workflow, take a look at our tech note and key presentation.

All our RNA sequencing workflows use NextSeq 500 instruments (Illumina).

Data processing and data analysis

RNA seq data processing

Processing of RNA sequencing reads is done by Cobra, our proprietary, automated and scalable platform. Cobra processes both small RNA as well as messenger RNA and long non-coding RNA sequencing data with state-of-the-art tools. Small RNA data processing is built on proprietary code based on a long standing expertise in the field as supported by our key publications in miRNA research. Cobra incorporates multiple quality control steps along the data processing pipeline. Data is quickly processed (more than 100 samples a day), while providing data security (against data loss and security breaches).

RNA seq data analysis

Basic data analysis includes differential expression analysis for small RNA, coding and long non-coding genes or transcripts. In addition, we offer comprehensive and customized data analyses such as benchmark studies, complex experimental design, pathway analysis, disease subtyping, among others. A PhD level data analysis expert is part of each project team.

Biomarker signature establishment

Model building for biomarker discovery is done in partnership with experts in the field. First, a feasibility study is done to evaluate the potential of your data to identify useful biomarkers. During this study, experts will provide a first estimation of the predictive performance of RNA sequencing results.

If the feasibility study outcome is successful, we proceed to biomarker signature establishment. Advanced mathematical model building also includes clinical information. A dedicated analysis selects promising candidate biomarker sets and provides a better understanding of the possible sensitivity and specificity trade-offs.

Dedicated resources to study non-coding RNA

Our data analysis pipelines include access to dedicated long non-coding RNA and small RNA functional annotation databases (such as decodeRNA and our proprietary LNCarta). We have generated an expanded human gene catalog and assembled a comprehensive reference transcriptome including lncRNA genes from dedicated databases such as LNCipedia and Ensembl.

LNCarta is Biogazelle’s proprietary database of predicted lncRNA functions through high-throughput perturbation by chemical compounds and silencing of transcription factors through siRNA. LNCarta assists in (1) mapping lncRNAs onto pathways, (2) providing functional context for lncRNAs, (3) and identifying upstream regulators of lncRNAs.