Aligners
Aligners softwares¶
The main alignment softwares are currently:
- BWA
- Bowtie
- STAR
They are all based on the Burrows-Wheeler Algorithm. This implies to build a genome index in which the genome is recoded using the BWA, ensuring very fast read alignments.
BWA-based aligner are CPU- and IO-demanding. In contrast they usually are not demanding in RAM (with maybe the exception of STAR, for index building)
Aligners take FASTQ (FASTQ.gz) filesas well as a genome reference index appropriately built as inputs.
They return BAM files which are compressed SAM files (Simple Alignment/Map).
The SAM format is really at the heart of RNAseq analyses, because it contains all the information needed to profile gene expressions from sequencing datasets.
Therefore, we highly recommend to take a few hours to look at all the details of the SAM format, which can be found in the GitHub repository. You can start with Sequence Alignment/Map format specification, and also have a closer look at Sequence Alignment/Map optional fields specification
Pseudo-aligners¶
Other aligners rather operate using a pseudo-alignment mode based on graphs of k-mers.