Test data - Drosophila melanogaster

[gbdias]

• Drosophila melanogaster is an interesting subject for test data since
  • the genome is small (±180Mb)
  • there are only 4 chromosomes (can expect 6 chromosome arms after scaffolding because the centromeres are hard to go through even with Hi-C)
  • plenty of sequencing data is available
• There is good PacBio HiFi data from the F1 of a cross between two highly inbred strains (A4 and ISO1).
• This should be a good heterozygous genome that assembles very well with hifiasm. I am currently testing the assembly to make sure. Will post results to this thread.
• There is Hi-C data for other strains, need to test if it works well.

[gbdias]

HiFi data

• https://ncbi.nlm.nih.gov/sra/SRX6957826[accn]
• This data was generated by PacBio for demonstrating Sequel and HiFi.
• It has an excellent read size (size selected to 24kb).
• The full dataset is 25.8 Gbp which corresponds to ~143x theoretical coverage.
• I was able to subsample 10% of the reads and still get a reasonable assembly in 13 minutes using hifiasm with 20cpus.
• Phasing is probably not great at this coverage but the haplotypes are similar enough that the final primary assembly is still very contiguous.

x-axis D.mel reference, y-axis hifiasm

[MartinPippel]

Is the subsampled data set already publicly available somewhere?