q2-sidle Error in docs

[lam (by)]

Missing option '--i-kmer-map'.
This document seems to have a mistake on https://q2-sidle.readthedocs.io/en/latest
when l do “ Reconstructing the Phylogenetic Tree”
There is no one option --i-regional-alignment but '--i-kmer-map which one is right？
I would like to ask the length of the kmer how to set , Such as V1-V2 Length is 241 Do I need to set the length of kmer and trim-length to 200
other region set 130
I have already tried this and no error has been reported but I don't know if that's true
sorry I am a beginner

[jwdebelius (Justine Debelius)]

Hi @lam,

Welcome to the :qiime2: forum!

@lam

Missing option '--i-kmer-map'.
This document seems to have a mistake on Welcome to Sidle — q2-sidle 2020.8 documentation
when l do “ Reconstructing the Phylogenetic Tree”
There is no one option --i-regional-alignment but '--i-kmer-map which one is right？

Thank you for finding that! We've had several people go through the docs, but apparently it was misssed; I will update the tutorial. The correct flag is --kmer-map.

@lam

I would like to ask the length of the kmer how to set , Such as V1-V2 Length is 241 Do I need to set the length of kmer and trim-length to 200
other region set 130

This depends on your database, region length, etc. There's no single correct answer and without having details about your specific primers, tables, etc, I don't know. What does your denoising look like? How long are those amplicons?

Finally, a business issue: please double check the forum code of conduct and don't double post or DM people. Public questions mean multiple people can find answers; cross posting can waste developer's time. We have a pretty good ecosystem for trying to make sure people get the help they need, but most mods are also full time researchers and are here on a volunteer basis in addition to our other work.

Best,
Justine

[lam (by)]

Thank you so much for your reply.
my primers find in this article Combining 16S rRNA gene variable regions enables high-resolution microbial community profiling
Table 1
Region # amplified Forward primer Reverse primer
Sequence Position Sequence Position
1 196415 5′-TGGCGGACGGGTGAGTAA-3′ 74 5′- CTGCTGCCTCCCGTAGGA-3′ 315
2 1122118 5′-TCCTACGGGAGGCAGCAG-3′ 316 5′- TATTACCGCGGCTGCTGG-3′ 484
3 660912 5′-CAGCAGCCGCGGTAATAC-3′ 486 5′- CGCATTTCACCGCTACAC-3′ 650
4 660342 5′-AGGATTAGATACCCTGGT-3′ 752 5′- GAATTAAACCACATGCTC-3′ 911
5 591604 5′-GCACAAGCGGTGGAGCAT-3′ 901 5′- CGCTCGTTGCGGGACTTA-3′ 1057
6 783882 5′-AGGAAGGTGGGGATGACG-3′ 1143 5′- CCCGGGAACGTATTCACC-3′ 1336

example find in github SMURF GitHub - NoamShental/SMURF

region rep-seqs info
region5 f901 r1057 length 156
|Sequence Count|Min Length|Max Length|Mean Length|Range|Standard Deviation|
|128 |135 |147 |140.6 |12 |2.0|
region1 F74 R315 length 241
|Sequence Count|Min Length|Max Length|Mean Length|Range|Standard Deviation|
|209 |205 |239 |223.29 |34 |6.27|

the two region as example，I've set region1 database kmer length and sidle trim-dada2-posthoc trim length 200
and region5 set length 130
I don't need to be specific, but I would like to know if this is the way to do it
should l set different length like this? Although sidle didn't report error when l do the reconstruction
and which output repseqs can directly used to do PICRUSt2 like QIIME, l can't finded

sorry I have just started learning

[jwdebelius (Justine Debelius)]

Hi @lam,

When I ran those, I think these are the parameters i used, based on the exaample from the SMURF repo

74f_315r:
  fwd: TGGCGGACGGGTGAGTAA
  rev: CTGCTGCCTCCCGTAGGA
  length: 215
316f_484r:
  fwd: TCCTACGGGAGGCAGCAG
  rev: TATTACCGCGGCTGCTGG
  length: 125
486f_650r:
  fwd: CAGCAGCCGCGGTAATAC
  rev: CGCATTTCACCGCTACAC
  length: 125
752f_911r:
  fwd: AGGATTAGATACCCTGGT
  rev: GAATTAAACCACATGCTC
  length: 125
901f_1057r:
  fwd: GCACAAGCGGTGGAGCAT
  rev: CGCTCGTTGCGGGACTTA
  length: 125
1143f_1336r:
  fwd: AGGAAGGTGGGGATGACG
  rev: CCCGGGAACGTATTCACC
  length: 175

@lam

which output repseqs can directly used to do PICRUSt2 like QIIME, l can't finded

Sidle doesn't integrate with PICRUSt - different algorithms, different approaches. I'd pick one amplicon - probably 486-650 and process through PICRUst separately.

Best,
Justine

[lam (by)]

thank you for your work very much！
That's what I was thinking
but there is a problem，only use one region to seems like will discard some asvs

[jwdebelius (Justine Debelius)]

Hi @lam,

You will, but the algorithms also aren't designed to handle multiple regions.

I might also personally consider whether you need PICRUSt at this point - I tend to find mixed reactions in the broader community and I might see what your data looks like before diving in there.

Best,
Justine