-
Notifications
You must be signed in to change notification settings - Fork 1
Input: How is my input supposed to be formatted or formulated?
marabouboy edited this page Dec 13, 2021
·
5 revisions
I will break down the input into four categories depending on how far you have proceeded in the bioinformatical analysis post-sequencing with either a MinION, a GridION, or a PromethION.
1. Fast5/.f5 format:
- In the case where you have just initialized the sequencing and your starting point is with the Fast5/.f5 sequencing format, my recommendation is to basecall the file and convert the file into Fastq format.
- This is simply done by using the guppy software, or any other open-source or proprietary basecalling software, according to the developer's specifications and instructions.
- Due to the Fast5/.f5 file being compressed as a default, I would only be able to show compressed files, therefore the figure is omitted.
2. Fastq/.fq format:
- If you've basecalled and/or are starting with the raw Fastq/.fq sequencing format, my recommendation is to align the sequences and convert the file into .bam format.
- This is easiest done with an established aligner, such as Minimap2, preferably with these three parameters
- Splice preset:
-x splice
- Samtools output format:
-a
- Removal of secondary alignment:
--secondary=no
- Complete command:
minimap2 -ax splice --secondary=no [Input.fastq] > [Output.sam]
- Splice preset:
- Ideally, your Fastq/.fq input file should have a formating similar to this:
2.5. SAM/.sam format:
- If you've aligned and/or are starting with the processed BAM/.bam aligned sequencing format, my recommendation is to sort and compress the sam-file into a bam-file. You can also optionally remove secondary reads and supplementary alignment as these reads may skew the results so as to enrich certain splice variants.
3. BAM/.bam format:
- If you've aligned and/or are starting with the processed BAM/.bam aligned sequencing format, my recommendation is to convert the bam-file into a BED12/.bed12 format using bedtools, as this BED12-file is the main input for FLAME.
- According to the developers of bedtools, the transformation into BED12 format requires the use of the bamtobed function and the bed12 flag.
- bamtobed function:
bedtools bamtobed
- BED12 flag:
-bed12
- Complete command:
bedtools bamtobed -bed12 -i [Input.bam] > [Output.bed]
- bamtobed function:
- Ideally, your BAM/.bam input file should have a formating similar to this:
4. BED12/.bed format:
- If you've already converted the input file into the prerequisite bed12 format, the input file is ready for FLAME
- My recommendation is to first test the pipeline by running the program without any additional input files. Refer to the wiki-section regarding reference creation and annotation.
- FLAME function:
./FLAME -I [Input.bed] -GTF [Reference.gtf]
- FLAME function:
- Ideally, your BED12/.bed input file should have a formating similar to this: