trying to build a new eukaryotic species SingleM metapackage

[fanch1122]

I am attempting to build a custom SingleM metapackage (smpkg) for eukaryotic species using BUSCO single-copy orthologs as marker genes. While the package construction completes successfully, all OTU annotations from short-read metagenomic data are returned as "root", indicating failed taxonomic assignment.

• Construction workflow

1. Marker selection: Extracted universal single-copy orthologs from BUSCO datasets
2. Sequence retrieval: Retrieved reference sequences for target eukaryotic clades
3. Taxonomy integration: Built custom NCBI taxonomy tree incorporating new lineages
4. Package assembly: Generated smpkg using singlem metapackage create

• Questions

1. Does SingleM officially support custom eukaryotic metapackages, or is eukaryotic analysis currently restricted?
2. Are there known limitations with short-read alignment against BUSCO-derived markers?
3. What diagnostic steps would help identify whether this is a taxonomy formatting issue vs. others failure?

• example otu_table.tsv

gene sample sequence num_hits coverage taxonomy
s3.108097 SRR12711264_R1 ACCGGCATCAAGGCCATTGACGGCATGATCCCCATCGGCAAGGGTCAGCGTGAGCTGATC 2 3.30 Root
s3.108097 SRR12711264_R1 ACAGGTATTAAGGCAATTGATGCCATGGTTCCAATCGGAAGAGGTCAGAGAGAGTTAATT 3 4.95 Root
s3.108097 SRR12711264_R1 ACCGGTATTAAATGTATCGACGCTCTCGTACCTATCGGACGTGGCCAACGTGAACTTATC 4 6.59 Root
s3.108097 SRR12711264_R1 ACCGGTATCAAGGTTGTTGACCTGATCTGCCCCTACGCAAAGGGCGGTAAGATCGGTCTG 3 4.95 Root
s3.108097 SRR12711264_R1 ACAGGCATAAAGGTGATTGACCTGCTGGAACCATACTGCAAAGGTGGGAAGATTGGACTC 1 1.65 Root
s3.108097 SRR12711264_R1 ACCGGCTTTAAGGCTATCGACGCGATGATTCCTATCGGTCGTGGTCAGCGTGAGTTGATT 6 9.89 Root
s3.108097 SRR12711264_R1 ACAGGCATTAAGGTAATAGATTTGCTCGAGCCCTACCTTAAAGGCGGCAAGATCGGTCTT 18 29.67 Root
s3.108097 SRR12711264_R1 ACAGGTATCAAGGCTATTGACAGTATGATTCCTATCGGCAGAGGCCAGAGAGAACTTATC 1 1.65 Root
s3.108097 SRR12711264_R1 ACCGGCATCAAGGCCATCGACTCCATGATCCCCATCGGTCGTGGCCAGCGTGAGCTGATC 2 3.30 Root
s3.108097 SRR12711264_R1 ACGGGCATCAAGGTCATCGATCTGCTCGAACCATATCTGAAAGGAGGAAAGATCGGACTT 1 1.65 Root
s3.108097 SRR12711264_R1 ACGGGCATCAAGGTCATCGACCTGATCTGCCCCTACGCCAAGGGTGGCAAGATCGGCCTG 3 4.95 Root

best wish ~

part_log.txt

[wwood]

Hi,

Not a small task! I would suspect your issue has to do with how you are generating the singlem package, rather than how you are running pipe. Making a metapackage (and singlem packages) is rather a non-trivial thing to do. You could look at some previous efforts for this in the extras folder in the repo e.g. new_package_creation or lyrebird_metapackage_creation.

Separately, we (i.e. @thepatientwait ) is actually trying to generate a euk-singlem as well. It is only in draft state at the moment, but it does give taxonomy outputs.

Can I ask what the context is here? Where and what are you working on?
Thanks, ben

[fanch1122]

Yeah, so I sincerely admire the amount and outstanding work your team has done. I carefully listened to your suggestions, made some adjustments to the demo metapacket, and made some updates using existing self-tested paired-end sequencing data. Perhaps the results from pipe indicate that this is feasible?

• report
  kingdom: 0.00% 0 taxons
  PM INFO: phylum: 0.00% 0 taxons
  PM INFO: class: 0.00% 0 taxons
  PM INFO: order: 0.00% 0 taxons
  PM INFO: family: 0.00% 0 taxons
  PM INFO: genus: 100.00% 24 taxons
  PM INFO: species: 0.00% 0 taxons

• otu.table
  gene sample sequence num_hits coverage taxonomy
  s3.108097 GQ12-1_1 ACTGGACAAAGAGCGCAAGATACATTTTTCCCTGTTGCTATAGGTGGAACATCTGCTATG 1 1.65 Root; d__Eukaryota; p__Ciliophora; c__Litostomatea; o__Pleurostomatida; f__Amphileptidae; g__Amphileptus
  s3.108097 GQ12-1_1 ACTGGACAGAGGGTAGTCGATGGCTTCTTCTCCCTGGCTAAGGGCGGAACCGCAGCTATC 161 265.38 Root; d__Eukaryota; p__Ciliophora; c__Prostomatea; o__Prorodontida; f__Colepidae; g__
  s3.108097 GQ12-1_1 ACAGGACAGCGTGTCATAGATACTATGTTCCCGCTGGCAAAAGGAGGAACCGCAGCAGTA 213 351.10 Root; d__Eukaryota; p__Ciliophora
  s3.108097 GQ12-1_1 ACAGGGCAGCGAATTATTGACACCATGTTCCCAATTGCTAAAGGTGGAGTTGCAGCAGTG 260 428.57 Root; d__Eukaryota; p__Ciliophora; c__Litostomatea
  s3.108097 GQ12-1_1 ACTGGCATTCTAGTGGTGGATGCAATGGTCCCTATAGGTCGCGGCCAGAGGGAGCTAATA 465 766.48 Root; d__Eukaryota; p__Ciliophora; c__Oligohymenophorea; o__Sessilida; f__Vaginicolidae; g__Vaginicola
  s3.108097 GQ12-1_1 ACTGGTATCAAAGTCGTCGATCTTCTTGCCCCATATGCCAGAGGAGGAAAAATCGGTCTC 34 56.04 Root; d__Eukaryota; p__Ciliophora; c__Prostomatea; o__Prorodontida; f__Colepidae; g__Levicoleps
  s3.108097 GQ12-1_1 ACCGGTATAAAGGCAATAGATCTCTTAATTCCATACCCAAAAGGCGGGAAGATAGGGCTT 113 186.26 Root; d__Eukaryota; p__Ciliophora; c__Prostomatea; o__Prorodontida; f__Colepidae; g__Coleps
  ...
  s3.108097 GQ12-1_1 ACTGGCATTCTAGTGGTGGATGCAATGGTCCCTATAGGTCGTGGCCAGAGGGAGCTAATA 2 3.30 Root; d__Eukaryota; p__Ciliophora; c__Oligohymenophorea
  s3.108097 GQ12-1_1 ACAGGGTTGAAAGCCATTGATACACTTGTTCCTATTGGACGAGGACAGCGCGAACTGATC 12 19.78 Root; d__Eukaryota; p__Ciliophora; c__Litostomatea; o__Pleurostomatida; f__Litonotidae; g__Litonotus
  s3.108097 GQ12-1_1 ACAGGGATCAAAGTTATAGATTTACTTGCTCCATATGTTAAAGGGGGAAAGATTCAGCTA 2 3.30 Root; d__Eukaryota
  s3.108097 GQ12-1_1 ACAGGAATTAAAACTATAGATTGTTTAATACCAATAGGTAGAGGTCAAAGAGAACTTATT 7 11.54 Root; d__Eukaryota; p__Ciliophora; c__Oligohymenophorea; o__; f__; g__
  s3.108097 GQ12-1_1 ACCGGTATAAAGGCAATAGATCTCTTAAGTCCATACCCAAAAGGCGGGAAGATAGGGCTT 1 1.65 Root; d__Eukaryota; p__Tubulinea; c__Elardia; o__Leptomyxida; f__Leptomyxidae; g__Leptomyxa
  ........

• Currently, I am pursuing my PhD at UCAS, doing some work in bioinformatics and Microbiology.

best wish ~

[wwood]

Hi @fanch1122

Indeed, looks like you've made some progress, though it is a little hard to tell from here. One thing, you shouldn't see g__ in the output - ideally all the taxons that go into the metapackage should be fully specified down to the species level. You may also need to generate an OTU table for singlem query (we usually generate a draft package without an sdb, apply it back to the genomes, manually modify the OTU table to have the correct taxon strings and then feed that to create the final metapackage. It's a bit fiddly..

HTH,
ben