miRanda output meaning?

Question

[stijn.vandenbrande]

Hello,

After using the miRanda tool the first time, I've got output like this:

Score: 167.000000 Q:2 to 20 R:3340 to 3360 Align Len (18) (83.33%) (94.44%)

Could someone tell me how to interpret? Is the "score" the miRSVR score? But then what does all the rest mean?

Many thanks!

[Buffo]

In order to parse the output try:

grep -A 1 "Scores for this hit:" miranda_out.txt | sort | grep '>'

you will get:

>mirna_name      transcript_target        143.00  -22.87  2 18    339 357 16      75.00%  87.50%

And this is the header:

mirna Target  Score Energy-Kcal/Mol Query-Aln(start-end) Subjetct-Al(Start-End) Al-Len Subject-Identity Query-Identity

Finally how are going? did you try pictar o targetscans?

[stijn.vandenbrande]

Hi,

No, I'm currently still using miRanda but I might complement it by using other Tools. Anyway, if you do not use the default parameters and therefore add this function " -strict " in the command, this means that you require strict alignment in the seed region (position 2-8). We already identified some interesting targets.

How are you doing?

[Buffo]

I´m pretty sure that result`s header information is available in its manual, but why use miranda? its algorithm allows wobble pairing in the seed wich may result in very untrusted target prediction, read this: target prediction tools But please, tell me about your results, which specie are you studying?

[stijn.vandenbrande]

Thank you for the fast reply. I didn't do miRanda analysis on the miRNAs of interest, I'm rather just learning how to interpret the result using the examples (which were present in the miRanda download). I will go through the article, but do you have a better alternative then miRanda for miRNA target predictions? The species I am studying is Locusta migratoria.

[Buffo]

I`m analyzing miRNAS (from mouse) too, but before target prediction I´m identifying expressed (based on database info), potential new mirnas,etc, etc, but I have read that the best algorithms are TargetScanS and picTar, but it also depends on your special interest, I think that this is the reason why there is not a "standard" pipeline to analyze miRNA-seq, I expect next week can start with target prediction, we would keep in touch for advices!

[stijn.vandenbrande]

Interesting research! I'll go through the article when I have some time and might change from miRanda to another tool. Sure, good idea