Error while training model

[xibalbanus]

G'day,

First of all I would like to thank you for such a useful program. I have been able to run with using the pre-trained models without any issues. However, when I try training with my own selection of OrthoDB groups I get the following error:

python ~/apps/ogcleaner/bin/ogcleaner.py train --orthodb_fasta crustacea.universal.single.fasta --threads 10 --clean
Command: /home/vader/apps/ogcleaner/bin/ogcleaner.py train --orthodb_fasta crustacea.universal.single.fasta --threads 10 --clean
OrthoClean model training module version 0.1a
Aligner: /home/vader/apps/ogcleaner/lib/mafft_bin/mafft
Aligner args: --anysymbol
Created directory train_orthodb_groups_fasta
Created directory train_nh_groups_fasta
Created directory train_paml_configs
Created directory logs
Created directory train_cluster_alignments_homology
Created directory train_cluster_alignments_nh
Created directory train_paml_trees
Created directory train_evolved_seqs
Created directory train_featurized_clusters
Created directory train_aliscores_homology
Created directory train_aliscores_nh
Beginning...
Segregating the sequences into their orthogroups...Traceback (most recent call last):
File "/home/vader/apps/ogcleaner/bin/ogcleaner.py", line 1750, in
main( args )
File "/home/vader/apps/ogcleaner/bin/ogcleaner.py", line 1426, in main
generate_trained_model( args )
File "/home/vader/apps/ogcleaner/bin/ogcleaner.py", line 1190, in generate_trained_model
ortho_groups = segregate_orthodb_groups( args.orthodb_fasta, args.orthodb_groups_dir, args.og_field_pos )
File "/home/vader/apps/ogcleaner/bin/ogcleaner.py", line 294, in segregate_orthodb_groups
t_group = m.group( 1 )
AttributeError: 'NoneType' object has no attribute 'group'

Any ideas about how to solve this?

Thanks in advance,

J.

EDIT: I fixed this (shown below) but I ran into a new error...

[xibalbanus]

If it helps, this is what the fasta file looks like:

294128:002b15 "pub_gene_id":"HAZT011030", "pub_og_id":"EOG090Y0000", "og_name":"actin binding","level":6657
LAQYASSLGDAAHGLEERVRLWRDYEATYERILRWMGEAEASLRDFCSRNSLQQKQEQLDKYQLVFASRHALVVEFCSIADKLEQNLICSMNQHQGDLEQLTNDSCDLM

(sequence truncated for sake of post)

Fixed by changing line 277 of ogcleaner.py to:

regex = re.compile( "\"pub_og_id\"+\:\"(\S*)\"" )

Maybe the fasta header changed from previous orthodb versions?

EDIT:

Another error after featurizing clusters and during the actual training of the model:

		Featurizing of EOG090Y05A0_NH Complete
		Featurizing of EOG090Y00YT_NH Complete
		Featurizing of EOG090Y002F_NH Complete
		Featurizing of EOG090Y03ER_NH Complete
		Featurizing of EOG090Y054H_NH Complete
	Done featurizing clusters!
	Saving featurized data set to train_featurized_clusters/featurized_data.txt... Done!
	Pickling featurized data set to train_featurized_clusters/featurized_data.pickle... Done!
Training model
Traceback (most recent call last):
File "/home/vader/apps/ogcleaner/bin/ogcleaner.py", line 1750, in <module>
	main( args )
File "/home/vader/apps/ogcleaner/bin/ogcleaner.py", line 1426, in main
	generate_trained_model( args )
File "/home/vader/apps/ogcleaner/bin/ogcleaner.py", line 1310, in generate_trained_model
	model, scaler = create_trained_model( args.model, args.features, data, args.threads )
File "/home/vader/apps/ogcleaner/bin/ogcleaner.py", line 899, in create_trained_model
	ss.fit( sdata[ features ] )
File "/usr/local/lib/python2.7/dist-packages/sklearn/preprocessing/data.py", line 560, in fit
	return self.partial_fit(X, y)
File "/usr/local/lib/python2.7/dist-packages/sklearn/preprocessing/data.py", line 583, in partial_fit
	estimator=self, dtype=FLOAT_DTYPES)
File "/usr/local/lib/python2.7/dist-packages/sklearn/utils/validation.py", line 407, in check_array
	_assert_all_finite(array)
File "/usr/local/lib/python2.7/dist-packages/sklearn/utils/validation.py", line 58, in _assert_all_finite
	" or a value too large for %r." % X.dtype)
ValueError: Input contains NaN, infinity or a value too large for dtype('float64').

Any ideas?

[masakistan]

Hey, so sorry for the late reply. I hadn't set my personal github account to 'watching' so I missed this. I apologize for that.

If it would still be helpful for you, I'd love to help.

Could you tell me how you got your OrthoDB groups? From what I can tell you grabbed from Crustacea with the Phyloprofile Present in all species and Single-copy in all species.

Let me know if this is correct and I can try running it on my end to see what I come up with.