SQANTI: extensive characterization of long-read transcript sequences for quality control in full-length transcriptome...

Area of Science:

• Transcriptomics
• Bioinformatics
• Genomics

Background:

• High-throughput sequencing with long reads enables novel transcript discovery.
• Characterizing these novel variants requires advanced computational tools.
• Mammalian transcriptomes, even well-annotated ones, harbor undiscovered transcript diversity.

Purpose of the Study:

• To present SQANTI, an automated pipeline for classifying and assessing the quality of long-read transcripts.
• To evaluate the effectiveness of SQANTI in characterizing full-length transcriptomes.
• To develop strategies for filtering technical artifacts from novel transcript data.

Main Methods:

• Development and application of the SQANTI pipeline, utilizing 47 unique descriptors.
• Analysis of a mouse neuronal transcriptome using Pacific Biosciences (PacBio) long reads.
• Experimental validation using Polymerase Chain Reaction (PCR) to assess novel transcript accuracy.

Main Results:

• SQANTI effectively characterizes full-length transcriptomes and identifies novel variants.
• A significant portion of novel transcripts identified by previous methods are technical artifacts.
• Novel transcripts primarily involve new splice site combinations, impacting transcript quantification and proteogenomic detection.

Conclusions:

• SQANTI provides essential tools for quality-evaluated and curated full-length transcriptomes from long-read sequencing data.
• The pipeline aids in distinguishing genuine novel transcripts from sequencing artifacts.
• Understanding novel transcript impact is crucial for accurate transcriptomic and proteogenomic analyses.