iq-tree error about pangolin

[varunshamanna]

Thank you for the update, I downgraded the iqtree and its working perfectly. Thank you for you your time. You can close the issue. Thanking You, Regards, Varun Shamanna Bioinformatician, CRL-KIMS, [email protected] [phone]

…

On Tue, 28 Apr, 2020, 3:47 pm aineniamh, ***@***.***> wrote: Thanks Varun, this seems to be a difference between iqtree 1 and 2 I didn't realise existed. For the moment, I'll change the environment file to explicitly need iqtree 1 (I'm running 1.6.12). If you've already got the environment setup, if you conda uninstall iqtree and then conda install iqtree=1.6.12 that should solve this problem. I'll switch everything over to iqtree2 more long term when I look into the differences (last I checked iqtree2 wasn't on conda, so have been working from iqtree1). — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#39 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AKJK2JGUU6ANY47B5KC5LX3RO2UKTANCNFSM4MSQSQ2A> .

[jillianhammond]

I am running into the same issue :)

[aineniamh]

Okay, that's a new one!
Could you try checking iqtree -v and maybe let me know what happens when you run

iqtree -s out/temp/query_alignments/tax1tax.aln.fasta -bb 1000 
-au -alrt 1000 -m HKY 
-g /home/crl-kims/miniconda3/envs/pangolin-2/lib/python3.6/site-packages/pangolin-0.1.1_2020_04_27-py3.6.egg/pangolin/scripts/../data/anonymised.aln.fasta.treefile 
-o 'outgroup_A'

I've not been able to replicate the error on my side so a bit more info will help figure this out.
Is there anything unusual about the query fasta that has been input? Any chance it contains a header with an empty sequence line?

I'll work today on adding in some more informative error messages.

[varunshamanna]

Thank you for the reply, I am having iqtree 2.0.3 and I ran the command you sent

itree -s out/temp/query_alignments/tax1tax.aln.fasta -bb 1000 -au -alrt 1000 -m HKY -g /home/crl-kims/miniconda3/envs/pangolin-2/lib/python3.6/site-packages/pangolin-0.1.1_2020_04_27-py3.6.egg/pangolin/scripts/../data/anonymised.aln.fasta.treefile -o 'outgroup_A'

I got this output

'For AU test please specify the number of bootstrap replicates via -zb option'

So then I changed -bb to -zb, I got the tree output I tried using different file also but no luck Thanking You, Regards, Varun Shamanna

…

On Tue, Apr 28, 2020 at 3:19 PM aineniamh ***@***.***> wrote: Okay, that's a new one! Could you try checking iqtree -v and maybe let me know what happens when you run iqtree -s out/temp/query_alignments/tax1tax.aln.fasta -bb 1000 -au -alrt 1000 -m HKY -g /home/crl-kims/miniconda3/envs/pangolin-2/lib/python3.6/site-packages/pangolin-0.1.1_2020_04_27-py3.6.egg/pangolin/scripts/../data/anonymised.aln.fasta.treefile -o 'outgroup_A' I've not been able to replicate the error on my side so a bit more info will help figure this out. Is there anything unusual about the query fasta that has been input? Any chance it contains a header with an empty sequence line? I'll work today on adding in some more informative error messages. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#39 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AKJK2JGU4BXO4YDQ5JUNYIDRO2REHANCNFSM4MSQSQ2A> .

[aineniamh]

Thanks Varun, this seems to be a difference between iqtree 1 and 2 I didn't realise existed.
For the moment, I'll change the environment file to explicitly need iqtree 1 (I'm running 1.6.12). If you've already got the environment setup, if you conda uninstall iqtree and then conda install iqtree=1.6.12 that should solve this problem.

I'll switch everything over to iqtree2 more long term when I look into the differences (last I checked iqtree2 wasn't on conda, so have been working from iqtree1).

[aineniamh]

Let me know if this solves it and I will close the issue.

[jillianhammond]

Hey Aine, Thanks for your help. The software runs for me however some of the sequences with assigned lineages have a bootstrap value of 0% despite being marked as success. I have copy and pasted some of the output to show you. Do you have suggestions?

taxon	lineage	SH-alrt	UFbootstrap	status
nCoV-012	A.1	0	0	success
nCoV-015	A.1	0	0	success
nCoV-027	A.1	0	0	success
nCoV-036	A.1	100	78	success
nCoV-017	A.2	100	98	success
nCoV-018	A.2	100	98	success
nCoV-020	A.2	100	99	success

[aineniamh]

No problem Jillian. I only just added in the status column yesterday and it refers to whether the sequences have passed the N-content and length requirements for assignment attempt. I realise now that might be slightly confusing as it doesn't relate to the confidence of the assignment (which the alrt and bootstrap should do) just whether or not the sequence passed QC. I'll revise this to return something more clear.
Perhaps something like this?
Status
Passed qc
Fail

[jillianhammond]

Thanks again Aine! Yes I think that would be more clear! :) I have another question
I am finding from the output that about 50-70% of my sequences are returning a bootstrap of 0% and if I rerun pangolin on the same fasta file the sequences returning 0% change (the lineages returned are the same each time, however). I am even getting 0% on the reference genome. The sequences I'm using do not contain any N's and are close to full length. Is this expected or perhaps an error on my end? Here is a copy and paste of some of the output from three runs of the same sequences. Thank you again for your help

nCoV-012_A	A.1	0	0	success
nCoV-015_A	A.1	0	0	success
nCoV-027_A	A.1	0	0	success
nCoV-036_A	A.1	0	0	success
nCoV-017_A	A.2	0	0	success
nCoV-018_A	A.2	0	0	success
nCoV-020_A	A.2	0	0	success
nCoV-022_A	A.2	100	99	success

taxon	lineage	SH-alrt	UFbootstrap	status
nCoV-012	A.1	0	0	success
nCoV-015	A.1	0	0	success
nCoV-027	A.1	0	0	success
nCoV-036	A.1	100	78	success
nCoV-017	A.2	100	98	success
nCoV-018	A.2	100	98	success
nCoV-020	A.2	100	99	success
nCoV-022	A.2	0	0	success

taxon	lineage	SH-alrt	UFbootstrap	status
nCoV-012	A.1	0	0	success
nCoV-015	A.1	100	93	success
nCoV-027	A.1	0	0	success
nCoV-036	A.1	0	0	success
nCoV-017	A.2	100	99	success
nCoV-018	A.2	100	98	success
nCoV-020	A.2	0	0	success
nCoV-022	A.2	100	98	success

[aineniamh]

So this was a bug, thanks for catching it! When I updated the guide tree I added in the SH-alrt stat to that too. iqtree keeps these legacy alrt and bootstrap values hanging around (apparently it's a "feature"), and I hadn't accounted for that. I've fixed this now, so it should be back running as normal. Hope this helps, thanks for that!