pre-filtering before mapping?

[NanYu]

Hi,
We have some mated paired end sequences (50bp) to be mapped to the reference genome for SNP (and maybe indel) discovery.
The data I have are csfasta and qual files.
Could any one let me know if I need to do pre-filtering for the sequences before I use any software to map them?
If I use bowtie, I should remove the orphan reads (and maybe try to map the orphan reads using a different parameter set).
If I use BFAST, should I do the same?

If I do need to filter the sequences based on the quality score, what's the cut-off threshold people normally use? Average of Q10?
How to translate the quality score to the % error rate like the Phred score?

Thanks!
Nan

[fanyucai1]

csfasta_quality_filter.pl this script will be used about qulity contorl in solid.
you can find this script by google

[fenciso]

Hi,

I am on index step, but apparently there is an error:

In function "RGIndexLayoutCreate": Fatal Error[OutOfRange]. Message: Layout must begin with a one.

Could someone help me with this problem???

[NanYu]

Hi,

I am on index step, but apparently there is an error:

In function "RGIndexLayoutCreate": Fatal Error[OutOfRange]. Message: Layout must begin with a one.

Could someone help me with this problem???

you mean the indexing of the reference genome using bowtie?

Below is the command I use (assuming bowtie-build and all_reference.fa are in the same folder)

./bowtie-build -C -f all_reference.fa reference_Color

[NanYu]

csfasta_quality_filter.pl this script will be used about qulity contorl in solid.
you can find this script by google

Thanks!
I downloaded the program. It has many parameters to use for trimming. Can anyone tell me normally what value they use for filtering (if any)? Thanks!

1. num_colors_to_hard_trim
2. min_median_qv
3. max_bad_colors_in_first_ten
4. max_number_bad_colors
5. num_consec_colors_to_trim
6. trim_terminal_bad_colors
7. min_read_length

[fanyucai1]

solid not supply the parameters, choose them by yourself, i advice you can statistics raw data before using last script.

[NanYu]

solid not supply the parameters, choose them by yourself, i advice you can statistics raw data before using last script.

Thanks! I am new in the area and would like some advice on choosing the values for those parameters, such as: Minimum QV value for a single color.

As to the statistics, besides median/average quality score, what else should I look into?

Thanks!

[fanyucai1]

there is a paper called :Analysis of quality raw data of second generation sequencers with Quality Assessment Software. you can find it by google. it is very simple ,it wil help you .
some parameters contains: min \max\Q20\mean\median you should consider

[paolo.kunder]

I downloaded and executet csfasta_quality_filter.pl script on my SOLiD 5500 csfasta (and qual) data to trim a fixed number of colors. ( code below)

perl csfasta_quality_filter.pl -f F5.cs fasta -q F5.QV.qual -o 5_bases_trimmed_F5.csfasta --num_colors_to_hard_trim 5

I was wondering about the output file, from the manual (Filtered and trimmed reads are output in csfasta format to a user-specified filename). And where is my associated QV.qual file trimmed? Any ideas? I cannot supply TopHat or any other alignment software with a trimmed csfasta and a full lenght associated quality file..

[fanyucai1]

I downloaded and executet csfasta_quality_filter.pl script on my SOLiD 5500 csfasta (and qual) data to trim a fixed number of colors. ( code below)

perl csfasta_quality_filter.pl -f F5.cs fasta -q F5.QV.qual -o 5_bases_trimmed_F5.csfasta --num_colors_to_hard_trim 5

I was wondering about the output file, from the manual (Filtered and trimmed reads are output in csfasta format to a user-specified filename). And where is my associated QV.qual file trimmed? Any ideas? I cannot supply TopHat or any other alignment software with a trimmed csfasta and a full lenght associated quality file..

This script could not output the .qual file after quality-contorl ,you could choose it from raw file according csfasta file .