Rasmus Kirkegaard
@kirk3gaard.bsky.social
Staff scientist having fun with DNA seq and bioinformatics at #AlbertsenLAB
Reposted by Rasmus Kirkegaard
claude: I have no idea why we now have more unmapped reads than before after adding in mismatch scoring
me: did you reverse complement the genome sequence?
claude: i'm a dumbass
me: did you reverse complement the genome sequence?
claude: i'm a dumbass
February 7, 2026 at 8:59 AM
claude: I have no idea why we now have more unmapped reads than before after adding in mismatch scoring
me: did you reverse complement the genome sequence?
claude: i'm a dumbass
me: did you reverse complement the genome sequence?
claude: i'm a dumbass
En anden ting er effektiviteten af bilen. Hvor mange km/min tilføjer man 😁⚡🚗
January 31, 2026 at 8:06 AM
En anden ting er effektiviteten af bilen. Hvor mange km/min tilføjer man 😁⚡🚗
When you look up something and all the links are already purple 😉
January 31, 2026 at 7:00 AM
When you look up something and all the links are already purple 😉
Reposted by Rasmus Kirkegaard
I think Single-M was written with that use-case in mind. Haven't tested it personally so can't vouch for it, but likely worth a try.
www.nature.com/articles/s41...
www.nature.com/articles/s41...
Comprehensive taxonomic identification of microbial species in metagenomic data using SingleM and Sandpiper - Nature Biotechnology
Novel microbial species in metagenomes are identified using conserved regions within universal marker genes.
www.nature.com
January 28, 2026 at 9:41 PM
I think Single-M was written with that use-case in mind. Haven't tested it personally so can't vouch for it, but likely worth a try.
www.nature.com/articles/s41...
www.nature.com/articles/s41...
Well we still mostly run denovo assembly of everything. But people like sylph for speedy classification of reads github.com/bluenote-157...
GitHub - bluenote-1577/sylph: ultrafast taxonomic profiling and genome querying for metagenomic samples by abundance-corrected minhash.
ultrafast taxonomic profiling and genome querying for metagenomic samples by abundance-corrected minhash. - bluenote-1577/sylph
github.com
January 31, 2026 at 6:23 AM
Well we still mostly run denovo assembly of everything. But people like sylph for speedy classification of reads github.com/bluenote-157...
Which assembler do you use?
January 21, 2026 at 6:33 PM
Which assembler do you use?
Makes sense with the many mods still to be covered in RNA. How much data do you get from RNA runs these days?
January 21, 2026 at 7:34 AM
Makes sense with the many mods still to be covered in RNA. How much data do you get from RNA runs these days?
Yeah we go SUP+mods on everything
January 20, 2026 at 11:10 AM
Yeah we go SUP+mods on everything
We basecall with modification calls and consider uploading just the bam. The added benefit of raw data is limited and cannot justify the storage needs?
January 20, 2026 at 10:33 AM
We basecall with modification calls and consider uploading just the bam. The added benefit of raw data is limited and cannot justify the storage needs?
The workaround is to pack Pod5 and fastq in on tar.gz and upload that as nanopore rawdata. It is not pretty but it works. However, I am not sure storing Pod5 is worth it.
January 20, 2026 at 10:33 AM
The workaround is to pack Pod5 and fastq in on tar.gz and upload that as nanopore rawdata. It is not pretty but it works. However, I am not sure storing Pod5 is worth it.
Pore occupancy is generally high to get to these yield numbers.
January 15, 2026 at 12:01 PM
Pore occupancy is generally high to get to these yield numbers.
Yes. These are metagenomic samples after bead beating so pretty much the sweet spot for sequencing.
January 15, 2026 at 6:00 AM
Yes. These are metagenomic samples after bead beating so pretty much the sweet spot for sequencing.
These are powermax +lfb cleanup before doing regular lsk preps. Sediments have been doing great for us.
January 13, 2026 at 5:33 PM
These are powermax +lfb cleanup before doing regular lsk preps. Sediments have been doing great for us.
I assume it depends on who is pouring the pores onto the flowcells at the factory (randomness).
January 13, 2026 at 4:55 PM
I assume it depends on who is pouring the pores onto the flowcells at the factory (randomness).
Yeah I expect that most of them are busy making cat videos but hopefully some of them will be available for science
January 9, 2026 at 12:49 PM
Yeah I expect that most of them are busy making cat videos but hopefully some of them will be available for science