Randfol in miRDeep2

[ccstaats]

Hello all,
I am trying to get miRDeep 2 to work with my data.
Everything goes fine with mapping, collapsing, etc.
When I try to run the mirDeep2.pl using the script

perl mirDeep2.pl collapsed.fa Genome.fa collapsed_vs_Genome.arf none none none

the initial process work fine. Until Randfold! The pipeline stops in "computing randfold p-values". One nucleus stay at 100% activity.
How long is this process? Here, 2 h and still counting.
Thank you for your help.
Charley

[ccstaats]

Dear all.
I Just realized that the process is quite long. The data analysis (4 Gb of fasta files) runned in 3 hours in a I5 processor with 8gb RAM.
Thank you all.
Charley

[jay2008]

RNAfold is most time-consuming task in predicting miRNAs from deep sequencing. All current predicting tools miRDeep, miRTRAP, MIReNA, miRDeep* and etc. use similar RNA secondary structure prediction algorithm. but it is not necessary to use big memory for preducting RNA secondary structure.

[JonB]

Hi,

I'm having trouble making randfold 2.0. I have the SQUID libraries but get this message when making randfold:

-bash-4.1$ make
gcc -O3 -I.  -o randfold params.o energy_par.o fold.o fold_vars.o utils.o randfold.c -lm -lsquid 
randfold.c:24:18: error: squid.h: No such file or directory
randfold.c:26:23: error: sre_random.h: No such file or directory
randfold.c: In function 'main':
randfold.c:30: error: 'SQFILE' undeclared (first use in this function)
randfold.c:30: error: (Each undeclared identifier is reported only once
randfold.c:30: error: for each function it appears in.)
randfold.c:30: error: 'sqfp' undeclared (first use in this function)
randfold.c:38: error: 'SQINFO' undeclared (first use in this function)
randfold.c:38: error: expected ';' before 'sqinfo'
randfold.c:89: warning: incompatible implicit declaration of built-in function 'strcpy'
randfold.c:96: error: 'SQFILE_FASTA' undeclared (first use in this function)
randfold.c:100: error: 'sqinfo' undeclared (first use in this function)
randfold.c:103: warning: incompatible implicit declaration of built-in function 'malloc'
randfold.c:103: warning: incompatible implicit declaration of built-in function 'strlen'
randfold.c:136: warning: incompatible implicit declaration of built-in function 'fabs'
randfold.c:158: warning: incompatible implicit declaration of built-in function 'free'
make: *** [randfold] Error 1

I am quite a novise when it comes to these things. Should I perhaps configure the makefile?

[gurjeet]

rndfold in mirdeep 2

I recently done small RNA sequencing of my endocrine tumour samples. I got result analysis with Mirdeep2 . I am not master in bioinformatics so , i outsourced the analysis. The company have predicted some novel micrornas. They have used mirdeep2 software for predicting novel microRNA.
I am facing problem in selecting novel microRNAs from small RNA sequencing. Is it necessary to have high mirdeep2 score with high mature read count. Those sequence with high mirdeep2 score ( like 3168) & mature read count (456) have non significant RANDFOLD p value. I am confused in this part.
Basically , I want to validate the results using PCR. But i am facing alot of problem in selecting top 5 novel microRNAs. I read in mirdeep2 documentation in which the score cut was from -10 to 10.
Please guide me regarding mirdeep2 software and please help me its parameters for selecting novel microRNA.I will be very thankful to you.
thanks.
Regards