I want to have a list of miRNAs unique for humans, e.i. those expressed in humans but not in mouse.
I checked miRBase but could not find any way to download sequences and names of miRNAs unique for humans.
I want to align miRNAs expressed in humans (I downloaded them from miRBase) to miRNAs expressed in mouse (again downloaded from miRBase), but cant find any online tool for this. Could somebody help me?
I want to compare seqeunces. Well, there are miRNAs with the same IDs in humans and mice, usually their seqeunces are very similar if not the same, and one would think that filtering with IDs is an option,but usually does not mean always :) That's way I think it's safer to compare sequences (intersection with IDs I can do).
Did you tried a BLAST tool for FASTA files? I know that such a comparison is possible. There are also some web tools for such alignment, however they are based on BLAST too. You can try a miRNAminer for instance. There is a lot of parameters to adjust when comparing sequences.
Doing such analysis with on-line tools can be a pain in such case. If I understood correctly, you simply need to take all human miRNAs and align them over mouse miRNAs? This might be a great opportunity to learn some scripting then!
I believe a tools such as BLAST can be used for this, yet you could also use Muscle for such tasks, your preference. If using blast, simply create the database using ./makeblastdb command from your Unix terminal. Then align your mouse sequences over human sequences. You need to exclude the human sequences, which have very good alignment scores, which can be done using e.g. python or similar tools, where you:
read all human miRNAs
if miRNA ID was in well aligned from the previous step, exclude it
else write it to a new file.
At the end you should have only miRNAs expressed in human.
Hi,
do you want to compare only the IDs of humans and mice miRs or some other attributes?
I want to compare seqeunces. Well, there are miRNAs with the same IDs in humans and mice, usually their seqeunces are very similar if not the same, and one would think that filtering with IDs is an option,but usually does not mean always :) That's way I think it's safer to compare sequences (intersection with IDs I can do).
Did you tried a BLAST tool for FASTA files? I know that such a comparison is possible. There are also some web tools for such alignment, however they are based on BLAST too. You can try a miRNAminer for instance. There is a lot of parameters to adjust when comparing sequences.
CD-HIT-2D could be a better option than blast from two complete databases, then you can filter unique sequences.