GridION fastq ASSEMBLY

Question

[shira.zaltsman]

Hey friends! I have a large number of fastq files of COVID-19 from GridION sequencing, I want to produce FASTA files of the assembly results. because the size of the reads is large (about 800-1200 bases) TRINITY didn't work well for me. what other tools can I use to assembly these files??

BTW: I was able to produce BAM sorted files using MINIMAP2 but I need the assembly results.

thank you for your help

Answer 1

[GenoMax]

Use an assembler meant for long nanopore reads. Flye is one of the current favorites.

Since there are many strains available at NCBI you may be fine with aligning to a close reference and then generating a consensus sequence: Generating consensus sequence from bam file

Alignment/assembly are two different things but you are fine with using either in this case.

[shira.zaltsman]

thanks! I installed Flye, but it yields an error:

    flye --nano-corr /data/tom/CORONA/fastq/SRR11313278.fastq.gz --genome-size 30k --min-overlap 1000 --out-dir /data/tom/CORONA/test
[2020-04-05 17:26:26] INFO: Starting Flye 2.7-b1587
[2020-04-05 17:26:26] INFO: >>>STAGE: configure
[2020-04-05 17:26:26] INFO: Configuring run
[2020-04-05 17:26:26] INFO: Total read length: 48522
[2020-04-05 17:26:26] INFO: Input genome size: 30000
[2020-04-05 17:26:26] INFO: Estimated coverage: 1
[2020-04-05 17:26:26] WARNING: Expected read coverage is 1, the assembly is not guaranteed to be optimal in this setting. Are you sure that the genome size was entered correctly?
[2020-04-05 17:26:26] INFO: Reads N50/N90: 1552 / 973
[2020-04-05 17:26:26] INFO: Selected minimum overlap: 1000
[2020-04-05 17:26:26] INFO: Selected k-mer size: 17
[2020-04-05 17:26:26] INFO: >>>STAGE: assembly
[2020-04-05 17:26:26] INFO: Assembling disjointigs
[2020-04-05 17:26:26] INFO: Reading sequences
[2020-04-05 17:26:26] INFO: Generating solid k-mer index
[2020-04-05 17:26:46] INFO: Counting k-mers (1/2):
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
[2020-04-05 17:26:46] INFO: Counting k-mers (2/2):
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
[2020-04-05 17:26:46] INFO: Filling index table
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
[2020-04-05 17:26:46] WARNING: No overlaps found - unable to estimate parameters
[2020-04-05 17:26:46] INFO: Extending reads
[2020-04-05 17:26:46] WARNING: No overlaps found!
[2020-04-05 17:26:46] INFO: Overlap-based coverage: 0
[2020-04-05 17:26:46] INFO: Median overlap divergence: 0
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
[2020-04-05 17:26:46] INFO: Assembled 0 disjointigs
[2020-04-05 17:26:46] INFO: Generating sequence
[2020-04-05 17:26:46] ERROR: No disjointigs were assembled - please check if the read type and genome size parameters are correct
[2020-04-05 17:26:46] ERROR: Pipeline aborted

[shira.zaltsman]

I checked in BLAST to see if there is an overlap and there is

[shira.zaltsman]

[GenoMax]

Does not look like you have much coverage (at least in that file) if the log above is to be believed.

[2020-04-05 17:26:26] INFO: Total read length: 48522
[2020-04-05 17:26:26] INFO: Input genome size: 30000
[2020-04-05 17:26:26] INFO: Estimated coverage: 1

You may want go the alignment and call consensus route if you don't have enough data for assembly.

[shira.zaltsman]

look like Flye, didn't work for viruses. what tool is good for alignment? I need to get a sequence

[GenoMax]

It looks like you don't have enough data for assembly, not that flye does not work. Follow the link above, for getting a consensus sequence from the data you have aligned using minimap2.