Hi,

It is very simple, download the miRNAs of your species or all the miRNAs in miRBase. And extract the first nucleotide. In UNIX/LINUX you can do it by running awk command. I assume miRNAs are in fasta format to extract the first nucleotide just run this command, grep -v ">" miRNAs.fa | awk -F "" '{print $1}' | sort | uniq -c. This command will give you the count/frequency of all the four nucleotides at first position. There you can use the data which ever way you want.

If you want to find the frequency of all four nucleotide at each position you have to iterate the command over full length of miRNAs.