A Result From Mummer Alignment

Question

[Love]

Hello, I run the mummer command for a pairwaise alignment

mummer -s -mum refer.txt query.txt > refer_query.align

The result is:

> query
1450      9370        21

gaggttgcagtgagctgagat

2122      9771        21

caaaaaaaaaaaaaaaaaaaa

3296      9570        26

gaggtcaggagatcgagaccatcctg

3318      9243        20

cctgaccaacatggtgaaac

3428      9353        30

cttgaacctgggaggcagaggttgcagtga

3461      9386        20

gagatcatgccactgcactc

3502      9773        20

aaaaaaaaaaaaaaaaaaag

My question is the result seems only display the matched strings in the query sequence. Can we also show the ones in the reference sequence?

[lh3]

If you are interested in variants, you can try dnadiff from MUMmer.

Answer 1

[Vitis]

I wonder what kind of alignment view you want or you need any specific format to parse with perl or python. The usual way I do nucleotide alignment with mummer is following:

align the two sequences

nucmer -p out seq1.fa seq2.fa

filter the results (delta format)

delta-filter -g -u 95 out.delta > out.fil.delta

show the alingment records (matching string coordinates)

show-coords -c -l -r -T out.fil.delta > out.fil.coords

show alignments for the regions you want (something like clustalw results)

show-aligns -r .....

[Love]

What I want to find at least say at least 80% identical match in reference and query. I need their positions in both sequence and letters. For example,

AGCTG 51 55 AGTTG 78 82

They have 4 same letters in 5. 4/5=80%

[Vitis]

I think you may need to filter based on identity at delta-filter part then use 'show-coords' to find the coordinates, then fetch the sequences from seq1 and seq2. There are a lot more stuff in the manual that I can't remember but I'm sure you'll find solutions there.

[Love]

Yes, I found delta-filter. What is meant by the header.
such as:

tagA1 tagB1 500 20000000

[Love]

I might spoke it correctly. Do you mean I need to create a delta file first? My question is that because nucmer uses mummer for its maximal exact matching, can I also find non-match?Is it an align file? How to generate it?

[Love]

I might spoke it incorrectly. Do you mean I need to create a delta file first? My question is that because nucmer uses mummer for its maximal exact matching, can I also find non-match?Is it an align file? How to generate it?

[Love]

I used the command nucmer -p out seq1.fa seq2.fa to align the two sequences, but why the align file is empty?

[Vitis]

any error message? You should see a file named out.delta, it's empty?

[Love]

The file exists but is empty(0kb). Do seq1.fa and seq2.fa need same length?

[Vitis]

that's quite strange, they don't have to be the same length. Are you sure they align with each other?

[Love]

I upload files here. Could you help me to look at them?

[Love]

I uploaded files here. Could you please help me to look at them?

[Vitis]

simply run 'nucmer refer.txt query.txt' I got a 'out.delta' and the alignment information. I'm working on RedHat 64 bit with the most recent version of mummer. The output is like this:

NUCMER 91432805 20000 20000 3232 3487 9156 9412 44 44 0 33 -32 -1 -55 8 0

[Vitis]

Any way to take this conversation to private emails? I may be able to help you troubleshoot the problem.

[Love]

Yes, it is a good idea. How can I find your email from the profile?