make and trainGlimmerHMM

[Llopis]

I have a problem to train my data with GlimmerHMM. The manual says:

Train GlimmerHMM module.

To use this trainning module, first run make in this directory, and than call
trainGlimmerHMM with the parameters specified below.

Usage:
trainGlimmerHMM <mfasta_file> <exon_file> [optional_parameters]

But when I run make in this directory (I am not the administrator but I have rights to execute and read), I get this:

make: Nothing to be done for `all'.

I thought that was everything in order but when I run my command line:

trainGlimmerHMM sequence.mfasta trainingdata.txt

I receive:

trainGlimmerHMM: command not found

I saw that other people already manage to train with their data. Could someone explain how you did it? Or how to solve this?

I think that something is wrong in the makefile:

# Copyright (c) 2003 by Mihaela Pertea.

# C compiler

C = gcc
CC = g++
CFLAGS = -O1 ${SEARCHDIRS}
#CFLAGS = -O3 -g -Wall
LIBS = -lm

MAKEFILE= makefile

.PHONY : all
all: build-icm build-icm-noframe build1 build2 falsecomp findsites karlin score score2 scoreATG scoreATG2 scoreSTOP scoreSTOP2 erfapp splicescore

How have you the makefile?

Thanks!

[TiborNagy]

If you do not have write permission, you can not compile the program.

[GenoMax]

If the trainGlimmerHMM exists in the directory you are trying to run it from, the file may need to have execute bit set before you can run the program. You *may* be able to do that by doing the following (or you may have to ask an administrator):

$ chmod u+x trainGlimmerHMM

Then you can run the program like this (assuming all these files in the current directory otherwise provide full file paths)

$ ./trainGlimmerHMM sequence.mfasta trainingdata.txt

If you are in doubt then post the long directory listing for trainGlimmerHMM file so we can see the permissions:

$ ls -l trainGlimmerHMM

[Llopis]

Oh, I see, so the problem it was not the make but my permissions. I solved that (Thanks TiborNagy and GenoMax) I run the program with:

perl ./trainGlimmerHMM sequence.mfasta trainingdata.txt

But I get the following message :

sh: line 1: 26107 Segmentation fault /bi/opt/GlimmerHMM/train/score train.acc train.facc train.don train.fdon score.acc score.don 0 0 1 > res.temp
ERROR 69: /bi/opt/GlimmerHMM/train/score exited funny: 35584 at trainGlimmerHMM line 445.

How can I arrange that?

[GenoMax]

Are the files you are using in the right format?

[Llopis]

The type I am using this exon data like this:

LD_Chr1_Draft_20130910_1349_consensus 585 3915
LD_Chr1_Draft_20130910_1349_consensus 4015 4301

LD_Chr1_Draft_20130910_1349_consensus 5605 5775

LD_Chr1_Draft_20130910_1349_consensus 6120 9450
LD_Chr1_Draft_20130910_1349_consensus 9550 9836

Note the inversion of the genes in the complement sequence

This exon file comes from a NCBI gene features file in fasta format, I am not sure that I did well the format conversion:

>lcl|NC_001144.5_cdsid_NP_013033.1 [gene=YLL067C] [protein=Putative Y' element ATP-dependent helicase] [protein_id=NP_013033.1] [location=complement(join(585..3915,4015..4301))]
>lcl|NC_001144.5_cdsid_NP_878115.1 [gene=YLL066W-B] [protein=hypothetical protein; overexpression causes a cell cycle delay or arrest] [protein_id=NP_878115.1] [location=5605..5775]
>lcl|NC_001144.5_cdsid_NP_013034.1 [gene=YLL066C] [protein=Putative Y' element ATP-dependent helicase; YLL066C is not an essential gene] [protein_id=NP_013034.1] [location=complement(join(6120..9450,9550..9836))]

Should be that like the first one or like this one?

LD_Chr1_Draft_20130910_1349_consensus 4301 4015
LD_Chr1_Draft_20130910_1349_consensus 3915 585

LD_Chr1_Draft_20130910_1349_consensus 5605 5775

LD_Chr1_Draft_20130910_1349_consensus 9836 9550
LD_Chr1_Draft_20130910_1349_consensus 9450 6120

EDIT: But with then I get this error: ERROR 36: no donors for training.
And this sequence (It is not a multi fasta, but I tried with adding a new sequence and it didn't work.)

>LD_Chr1_Draft_20130910_1349_consensus
TAAAGTTATCCACAGCTTGTGGACAGTTTTAATTTTATTTTGATAAGCCCTGTTAACACAACGTTTAGTTATCCACAGAG
CGCTGTGAATTCCTTGAATTAAACTTTTGTTTTTCCACTTATCCACAGAAGAGTTATCCCAACTGTCTACAACTGTGAAA
CAATTATGCACAGGCTTGTTTTGCCTGTGGAAAACTTTTATTGAGAATGCTAGAATTAATTTGCATTTCATAAAACGACT
AAAAATCAGGGGGGAATTTCAAAGTGCTTGATAAGGATTCATTGTGGCAAAAACTAAGTGAACAATTTCGAGAGAACACA...

I see it starts to work because it prints this:

-bash-4.1$ perl ./trainGlimmerHMM ~/LD_Chr1_Draft_20130910_1349_Cons_R.fasta ~/trainingdata.txt
Simple Consensus = aaacaatcatattaatatgaaaatagaaa
Markov Consensus = atatgccaaaactgccatgaaagtagcca
******** Old Way = aactaagctgatcaatatgaaaaaagata
Simple Consensus = taattattataaaaatttttaataagtta
Markov Consensus = aaaaaaaaaaaaaaaaaaaaaaaaagaaa
******** Old Way = ttttttttttttttttttttttttagttt
Simple Consensus = acggcgttcaagcctt
Markov Consensus = acggcgttcaagcctt
******** Old Way = acggcgttcaagcctt
Simple Consensus = aaatggttattaaaat
Markov Consensus = aattggtttttttttt
******** Old Way = aaatggtttttttttt
Consens 0 pe poz 20
Frunze de 8 si 0 componente
Consens 0 pe poz 0
Frunze de 2 si 0 componente
rm: cannot remove `outf*': No such file or directory
Simple Consensus = aaataacgtaacgcaacaaaaagaactgaggacttatcttaaataattgggtaaaaagagattataactattacaactac
Markov Consensus = gatttagataagttaatagttcatctcgagaagctgccgacatgagaagggtaattcggcttttcctgttaacttagaat
******** Old Way = ataaaacccgacccaataaaaagaacgtgtgagctgccgacatgagaagggtaaaaaaagattaggacaattacaacgtc
Simple Consensus = ttatttaataaatttataattattattataaatattaattttataattattattaattttattattattattttaatttt
Markov Consensus = tttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttt
******** Old Way = tttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttt
Simple Consensus = ctaaggtggcttatgcccaatctaaagtcgcagatgcgttatttgccgttgaattgaataaacgggctgagacagacggc
Markov Consensus = ctaaggtggcttatgcccaatctaaagtcgcagatgcgttatttgccgttgaattgaataaacgggctgagacagacggc
******** Old Way = ctaaggtggcttatgcccaatctaaagtcgcagatgcgttatttgccgttgaattgaataaacgggctgagacagacggc
Simple Consensus = ttattattattattattattattattattattattattattattattattattataattattattattattattaaaatg
Markov Consensus = tttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttg
******** Old Way = tttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttg
Simple Consensus = gaattatataaaatgccaaagtgcaaaggatatcataaataattaatggaagttacaaacaatcatattaatatgaaaat
Markov Consensus = gcttcaaccagcctactggggttttttgtgtggaattgttttgcatttgacgtttattatgccaaaactgccatgaaagt
******** Old Way = gcttcaaccaaaacgacaaaccccaaaagagatcataaaccgttaattgaaagaccaactaagctgatcaatatgaaaaa
Simple Consensus = tattattaaaaatattatttttattaataattttattattttaattataattattaaaattattaataaaatttttaata
Markov Consensus = tttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttaaaaa
******** Old Way = tttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttt
Simple Consensus = tcaagcctttgctgttcacccgggattagttcccggaaccggcttgggacgttatacaacgcataatggggcagttcgac
Markov Consensus = tcaagcctttgctgttcacccgggattagttcccggaaccggcttgggacgttatacaacgcataatggggcagttcgac
******** Old Way = tcaagcctttgctgttcacccgggattagttcccggaaccggcttgggacgttatacaacgcataatggggcagttcgac
Simple Consensus = tattaaaattattattatttttattattattattattattattattattattattataattaatattattattattatta
Markov Consensus = tttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttt
******** Old Way = tttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttttt
sh: line 1: 53413 Segmentation fault /bi/opt/GlimmerHMM/train/score train.acc train.facc train.don train.fdon score.acc score.don 0 0 1 > res.temp
ERROR 69: /bi/opt/GlimmerHMM/train/score exited funny: 35584 at ./trainGlimmerHMM line 445.

[ersgupta]

Hi,

Were you able to resolve the "ERROR 69: segmentation fault"? I am facing the same problem.

[Llopis]

Hi ersgupta,

I think I did solve it, i'm sorry but I can't remember, . I couldn't find anything on my notes... sorry.

[ersgupta]

Hi Llopis,

No problem. I was able to finally resolve it, by making some array size adjustments to the code. thanks.