talanki wrote:
> Hi,
>> I am trying to assemble sequences using the phred,cross_match and phrap
> softwares. I read the manuals and am a little bit confused on how to
> use them.
>> I have 2 sequence files
> 1. abc_F.abi - forward read i.e read from 5' to 3' on the sense
> strand and the other is,
> 2. abc_R.abi - reverse read i.e read from 5' to 3' on the antisense
> strand.
>>>> I need to create a contig of these two sequences.
> These are the following steps I performed.
>> 1)
> I used phred and created the following files
>> abc_F.abi.seq
> abc_F.abi.qual
> abc_R.abi.seq
> abc_R.abi.qual
[...]
> My first doubt is how to assemble the above 2 sequences using phrap.
>> Should I write both the sequences in tht single file?
Yes. Also the quals should be in a single file. Actually phred can do
this for you with the -sa and -qa options. Eg:
phred -if my_input_traces_filelist.txt -sa abc.fasta -qa abc.fasta.qual
> What should be the naming convention of the file?
phrap doesn't care. But if your .qual file has the exact same name as
the sequence + a ".qual" suffix, then phrap will automatically use the
quality values to help it assemble. Otherwise the quality values will
not be used.
>> Can I get a brief description on how to use phrap. the phrap
> documentation is confusing and I am not able to understand.
>
phrap abc.fasta
phrap attempts to assemble all the sequences in "abc.fasta" using the
quality values in "abc.fasta.qual" if such a file exists. Among the
output files created will be a file of contig sequences and a file of
singlet (unassembled) sequences.
> Can anybody help me?
>
No need to do the vector cross_matching for just F/R reads from a single
clone. If you do want cross_match to help you identify vector, I would
suggest you run cross_match on the contig after assembly.
phrap is pretty high-powered--typically one uses it to assemble hundreds
or thousands of reads. But I do use it for simple F/R read assemblies.
www.phrap.org has detailed documentation for phred, phrap and consed.
But much of it requires fairly in-depth study to use if you are not
already familiar with the programs.
> bye,
> LT
>
Feel free to ask more questions if this doesn't work for you.
--
Phillip SanMiguel
Purdue Genomics Core Facility
