java.lang.RuntimeException: java.lang.StringIndexOutOfBoundsException: String index out of range: -1

[xiekunwhy]

Hi,

I got these errors and rna-bloom stuck forever, what's wrong? Exception in thread "Thread-16" Exception in thread "Thread-23" Exception in thread "Thread-21" Exception in thread "Thread-22" Exception in thread "Thread-20" Exception in thread "Thread-19" Exception in thread "Thread-17" Exception in thread "Thread-18" java.lang.RuntimeException: java.lang.StringIndexOutOfBoundsException: String index out of range: -1 at rnabloom.RNABloom$FastqToGraphWorker.run(RNABloom.java:617) at java.base/java.lang.Thread.run(Thread.java:834) Caused by: java.lang.StringIndexOutOfBoundsException: String index out of range: -1 at java.base/java.lang.StringLatin1.charAt(StringLatin1.java:47) at java.base/java.lang.String.charAt(String.java:693) at rnabloom.bloom.hash.NTHash.NTPC64(NTHash.java:471) at rnabloom.bloom.hash.NTHash.NTMC64(NTHash.java:702) at rnabloom.bloom.hash.CanonicalPairedNTHashIterator.next(CanonicalPairedNTHashIterator.java:41) at rnabloom.RNABloom$FastqToGraphWorker.run(RNABloom.java:574) ... 1 more java.lang.RuntimeException: java.lang.StringIndexOutOfBoundsException: String index out of range: -1 at rnabloom.RNABloom$FastqToGraphWorker.run(RNABloom.java:617) at java.base/java.lang.Thread.run(Thread.java:834) Caused by: java.lang.StringIndexOutOfBoundsException: String index out of range: -1 at java.base/java.lang.StringLatin1.charAt(StringLatin1.java:47) at java.base/java.lang.String.charAt(String.java:693) at rnabloom.bloom.hash.NTHash.NTPC64(NTHash.java:471) at rnabloom.bloom.hash.NTHash.NTMC64(NTHash.java:702) at rnabloom.bloom.hash.CanonicalPairedNTHashIterator.next(CanonicalPairedNTHashIterator.java:41) at rnabloom.RNABloom$FastqToGraphWorker.run(RNABloom.java:574) ... 1 more

Best, Kun

[kmnip]

Can you please report the version and the exact command you used?

[xiekunwhy]

Hi,

I use latest conda version, and jar v1.4.3 also has the same problem.

I assembled 30 samples separately, only 3 of them were finished normally, 27 samples got above errors.

here is the log file rna-bloom.zip

Best, Kun

[kmnip]

Thanks. I am posting part of your command here for future references:

rnabloom -sensitive true -bound 500000 -f true -left ZD_E28_3_1.fq.gz -right ZD_E28_3_2.fq.gz -revcomp-right -t 8 -outdir ZD_E28_3

There are several things wrong with your command:

1. -sensitive does not take any arguments. -sensitive true would not work.
2. -f does not take any arguments. -f true would not work. Typically, you don't need to use this option unless the command is a re-run and you want to overwrite all existing files.
3. -bound 500000 is set way too large. I would not expect the fragment size for your paired-end reads to be 500Kbp. Usually, it is well under 1,000 for a typical Illumina RNA-seq sample. Setting a bound that high will result in an extremely long runtime for stage 2.
4. Since you have installed from conda, please include the -ntcard option to automatically calculate the appropriate Bloom filter sizes. From one of the log files, I noticed that stage 1 was re-run automatically due to a very high Bloom filter false positive rate.

Taking all this together, this modified command should work:

rnabloom -ntcard -sensitive -bound 500 -f -left ZD_E28_3_1.fq.gz -right ZD_E28_3_2.fq.gz -revcomp-right -t 8 -outdir ZD_E28_3

[xiekunwhy]

Thank you for pointing out.

Sorry, the .sh files above is the second run, and the log files are the first run's log. here is the first run command line: rnabloom -left ZD_E28_3_1.fq.gz -right ZD_E28_3_2.fq.gz -revcomp-right -t 8 -outdir ZD_E28_3 all are default except input and output parameters.

I changed the parameters as I want to let rnabloom run through, but failed again.

Best, Kun

[kmnip]

What is the error this time? And, can you please remove the output directory ZD_E28_3 if it already exists and try again?

[kmnip]

Also, please use the ntcard option as mentioned previously.

[xiekunwhy]

Hi,

rnabloom finished normally this time (rnabloom -left ZD_E28_3_1.fq.gz -right ZD_E28_3_2.fq.gz -revcomp-right -t 20 -outdir ZD_E28_3 -ntcard), but produced too many transcripts (126327 transcript in rnabloom.transcripts.nr.fa file), the gene number of this species is about 20000, so 126327 transcript is too many. Is there a way or some parameters to reduce transcript number?

Best, Kun

[kmnip]

You may consider using EvidentialGene: http://arthropods.eugenes.org/EvidentialGene/about/EvidentialGene_trassembly_pipe.html

[xiekunwhy]

EvidentialGene is not so easy to use.