How To Make Statistics Of Mismatch And Unpairing Positions Between Mirnas And Their Targets

Question

[biolab]

Dear all,

i need to make statistics of mismatch and unpaired positions between miRNAs and their targets. What I have done is that I performed patscan and identified miRNA targets using various patterns ( allowing different mismatch and unpairing numbers ). My patscan result is examplified as below.

*An example only*
miRNA    miR_sequence   target_antisense_sequences
miR-1    AGCGCU         AGCTCT
miR-2    AGAGAG         AGAAG
mIR-3    UGAGC          TGAGGC

My problem is to make statistics of unparing and mismatch positions. To resolve this problem, I performed batch blast2 to compare miRNAs (query ) and their targets (subject), and output is like below.

*concise BLAST Result*

miR1
  AGCGCU 
  111 11
  AGCTCT


miR2
 AGAGAG  
 111 11
 AGA-AG


miR3
 UGAG-C  
 1111 1
 TGAGGC

I think it's close, but it's still troublesome to make statistics of mismatch and gap positions. My ideal result is shown below.

        mismatch_position    gap_position
miR1    4                    0
miR2    0                    4
miR3    0                    5

I think my analysis is a little bit different from fuzzy matching. My key problem is to calculate mismatch and unpairing positions between miRNAs and their targets. My blast result is close, but i don't know how to statistically summarize where mismatchs and gaps are for many miRNA-target pairs. Could anyone please help me sort this problem? THANK YOU VERY MUCH!