Unexisting tree file or Malformed newick tree structure

[aberaslop]

Hi! First thank you so much for such a useful program!

I have encountered an error when running avp detect, which is pasted below. I used diamond following your mention in the wiki, which worked well. Any pointer to what is going on would be extremely helpful.

Thanks!!
Linnie

~/software/AvP/avp detect -i C2_output/mafftgroups/ -o C2_output/ -g C2_output/groups.tsv -t C2_output/tmp/taxonomy_nexus.txt -c config.yaml
[+] Setting up
[!] Found 2617 groups and 4370 genes
[+] Reconstructing phylogenies with FastTree
[x] 100%
[+] Analyzing fasttree results
Traceback (most recent call last):
File "/scistor/guest/zrs382/software/AvP/avp", line 7, in
main()
File "/net/sys/pscst000/export/bazis/guest/zrs382/software/AvP/depot/interface.py", line 32, in main
detect.main()
File "/net/sys/pscst000/export/bazis/guest/zrs382/software/AvP/depot/detect.py", line 226, in main
make_nexus_file(gene, group, lineage, gene_nexus_path, group_file, phylogeny_file, colors)
File "/net/sys/pscst000/export/bazis/guest/zrs382/software/AvP/depot/PetNexus.py", line 60, in make_nexus_file
t = Tree(phylogeny_file)
^^^^^^^^^^^^^^^^^^^^
File "/scistor/guest/zrs382/.local/lib/python3.11/site-packages/ete3/coretype/tree.py", line 212, in init
read_newick(newick, root_node = self, format=format,
File "/scistor/guest/zrs382/.local/lib/python3.11/site-packages/ete3/parser/newick.py", line 264, in read_newick
raise NewickError('Unexisting tree file or Malformed newick tree structure.')
ete3.parser.newick.NewickError: Unexisting tree file or Malformed newick tree structure.
You may want to check other newick loading flags like 'format' or 'quoted_node_names'.

The output of avp detect was:

Setting up
[!] Selected 4370 HGT candidates
[+] Parsing hits file and grouping similar queries
[!] Formed 2617 groups
[+] Extracting hits from DB
[+] Writing fasta files
[!] Skipped 432098 hits and 0 taxids.
[+] Aligning fasta files
[x] 100%
[!] Finished with 4370 HGT candidates in 2617 groups

This is my config file:

max_threads: 4

DB path

blast_db_path: /scistor/guest/zrs382/databases/ # blast: change to the local blast_db path
fasta_path: /scistor/guest/zrs382/databases/ # diamond: change to the local fasta path for sp, ur90, or custom database
mode: blast # use blast for blast database, use sp for swissprot database, ur90 for uniref90 or custom database
data_type: AA # data type DNA, AA

Algorithm options

prepare

ai_cutoff: 0
ahs_cutoff: 0
outg_pct_cutoff: 80
selection: "ai or ahs" # select sequences based on which metrics, another example "(ai or ahs) and outg_pct"
percent_identity: 100 # select hits equal or below this number
cutoffextend: 20 # when ingroup hit is found, we take this hit + n hits
number_hits_noingroup: 50 # when no ingroup hit is found, we take this number of hits
trimal: false
min_num_hits: 4 # select queries with at least that many blast hits
percentage_similar_hits: 0.7 # group queries based on this

detect, clasify, evaluate

fastml: true # Use fasttree instead of IQTree
node_support: 0 # nodes below that number will collapse
complex_per_ingroup: 20 # if D/(D+I) smaller than this then node is considered Ingroup
complex_per_donor: 80 # if D/(D+I) greater than this then node is considered Donor
complex_per_node: 90 # if node contains percent number of this category, it is assigned

Program specific options

mafft_options: '--anysymbol --auto'
trimal_options: '-automated1'

#IQ-Tree
iqmodel: '-mset WAG,LG,JTT -AICc -mrate E,I,G,R'
ufbootstrap: 1000
iq_threads: 4

[GDKO]

Hi @aberaslop,

You should point to the nr with blast_db_path

For example
blast_db_path: /scistor/guest/zrs382/databases/nr
if nr is there, and rerun from avp prepare

Cheers,
Georgios

[aberaslop]

Hi, @GDKO,

Thank you so much for your prompt answer! I updated the path and re-run avp prepare, with the same result. If there is anything else you think may be going on, I'd be happy to try.

Thanks a lot!

Aileen

[GDKO]

Hi @aberaslop,

It appears that it cannot fetch the sequences from the database. Can you post the diamond command you used and the updated config file?

Cheers,
Georgios

[aberaslop]

Dear Georgios,
Thank you for your answer. I decided to run everything from scratch and it worked just fine after performing your adjustments. I have a couple of questions, one regarding the pipeline, another the results.

1. If I want to run AvP with diamond, using fasta_path for the fasta file used to build the diamond database, should I leave blast_db_path empty? And what should I fill out in mode? (All of this regarding the config.yaml file)

2. I ran AvP on my dataset and it revealed potential HGT from bacteria that were expected (previously published). However, I eventually realized that the run had crashed, so I re-run it again. After running successfully, those potential HGT are no longer detected, they have been now included in the Ingroup folder. I have blasted the non-target sequences in those ingroups, and they are all very close relatives of our target organism. Would you expect this behavior if the first run finished unexpectedly, and the HGT occurred in the common ancestor of the target species and the close relatives?

Thanks a lot!
L.

[GDKO]

Hi @aberaslop ,

1. Did you follow #11 for diamond and nr?
2. At what step did it crush? Also, how do the trees look like for these HGTs?

Cheers,
Georgios