[GenoMax]

Hi
I want to do WGS of 2 samples of human. I have 2 sequencer Hiseq2000 in main facility. I have experience of Analysing of Illumina data. But the problem is that I am not sure about the following:
1. replication / replicates
2. depth coverage
3. best stretegies
4. Kits / reagents

May I ask this forum, if you can give me guideline on the sequencing for Whole Genome. Since this field is developing very fast, I dont want to any mistake and asking here. I have funding and sequencer (freely available), just need to buy kits and reagents. Please tell me how can I design best for WGS.
Thank you, (my English is not good so please ignore English mistakes)
jp.

This calculator can help with the coverage calculations: http://support.illumina.com/download...alculator.ilmn
and a tech note to go with it: http://res.illumina.com/documents/pr...alculation.pdf

[jp.]

I thank you for providing kind valuable info. My samples are from human origin Paired-End-HiSeq2000 and I have to pay for reagents cost only because HiSeq2000 is just free to use, in my case, budget is not a problem.
Therefore, I like to ask few more question;
1. How many replicates are required for whole genome sequencing ?
I am thinking to go for 3 biological replicates + 3 technical replicates per sample i.e. 1 samples = (1 x 3) 3 = 9. Is it enough ? or I should increase the replicates ?
2. What is the best depth coverage, 30X or more ?
3. How can I increase depth coverage ?
4. What is best library size, 150 or 250 or higher?
5. Can I use single-cell DNA/RNA on Paired-End-HiSeq2000 ?
Thank you very much in advance
(I am sorry for asking very preliminary question here but nobody answers my question, every body want something back to give me the answers or just said, figure it out. There is nobody in my location to help/guide a beginner).

[Heisman]

I thank you for providing kind valuable info. My samples are from human origin Paired-End-HiSeq2000 and I have to pay for reagents cost only because HiSeq2000 is just free to use, in my case, budget is not a problem.
Therefore, I like to ask few more question;
1. How many replicates are required for whole genome sequencing ?
I am thinking to go for 3 biological replicates + 3 technical replicates per sample i.e. 1 samples = (1 x 3) 3 = 9. Is it enough ? or I should increase the replicates ?
2. What is the best depth coverage, 30X or more ?
3. How can I increase depth coverage ?
4. What is best library size, 150 or 250 or higher?
5. Can I use single-cell DNA/RNA on Paired-End-HiSeq2000 ?
Thank you very much in advance
(I am sorry for asking very preliminary question here but nobody answers my question, every body want something back to give me the answers or just said, figure it out. There is nobody in my location to help/guide a beginner).

You need to describe your experiment more fully to get valid answers. Different coverage is required for different applications. If you're trying to call SNPs as either heterozygous or homozygous, you need much less coverage than if you are trying to find mutations present at a frequency of say 5% in a tumor sample. Read length depends on your application too. For example, if you're calling mutations then all else being the same longer reads may be better as they will be easier to align, assuming the quality doesn't tail off too much at the ends. However, if you're doing expression analysis where it is the number of reads that matters moreso than the specific coverage of any given base, longer reads won't help as much (although they may help if you're looking for different isoforms). These might be the reasons people want you to give "something back" to get the answers.

[jp.]

Thank you so much for your valuable info. I have studies a bit and wish to ask few more things.
I wish to sequence whole genome to study SNPs, mutations. I will also use for expression study. I have to sequence transcriptom, So I will do RNA-seq for expression and mutation. I also want to compare the expression data from RNA-seq with Whole Genome Seq. I am thinking to compare this whole genome with hg19 and then compare RNA-seq with hg19 and my whole genome seq. Overall, I will do complete analysis. I also want to build my own reference whole genome which I will use as a hg19 and vice versa.
1. Do I have to go for two different time sequencing of whole genome with normal coverage and longer and shorter reads ?
2. I wonder how hg19 is built, the same way I want to built my own. Is it built by sanger.

Expecting kind reply.
Thank you.

You need to describe your experiment more fully to get valid answers. Different coverage is required for different applications. If you're trying to call SNPs as either heterozygous or homozygous, you need ......

[jp.]

Thank you Heisman
Finally, I am going with 500 reads with >30X WGS. Is there any suggestion ?

You need to describe your experiment more fully to get valid answers. Different coverage is required for different applications. If you're trying to call SNPs as either heterozygous or homozygous, you need much less coverage than if you are trying to find mutations present at a frequency of say 5% in a tumor sample. Read length depends on your application too. For example, if you're calling mutations then all else being the same longer reads may be better as they will be easier to align, assuming the quality doesn't tail off too much at the ends. However, if you're doing expression analysis where it is the number of reads that matters moreso than the specific coverage of any given base, longer reads won't help as much (although they may help if you're looking for different isoforms). These might be the reasons people want you to give "something back" to get the answers.

[Heisman]

Sorry I did not respond earlier. I am confused. Why do you want to build your own reference genome? What do you mean in your last post when you say you are going for 500 reads? You seem to want to do a lot of different experiments; do you have people around you who you can further discuss this with?

[jp.]

Thank Heisman
As I wrote earlier, I am almost a lone wolf :|
However, I have changed my plans.
Now want to detect clinically relevant Variants and might look something common between my my available RNA-seq data and thinking for 500 insert PE.
Can you please tell me your best strategies, if you were at my place ?
Thank you

Sorry I did not respond earlier. I am confused. Why do you want to build your own reference genome? What do you mean in your last post when you say you are going for 500 reads? You seem to want to do a lot of different experiments; do you have people around you who you can further discuss this with?

[Heisman]

I haven't done RNA-seq, hopefully someone else can chime in. Why do you want 500 np inserts?

[jp.]

because I think it will be good to detect variants and mutation. Also, I think that this much will be good enough to study SNP and gene expression.
What are your recommendations for some one who is doing WGS for the first time ?
what should one study if checking control vs treatment or normal vs diseased ?
How much should be insert and depth ?
May I know your opinion, if your guide gives you human samples (control vs treatment, and normal vs diseased) and if you are doing first time WGS ?
You may please write which may not be perfect but at least to start WGS with ?

I haven't done RNA-seq, hopefully someone else can chime in. Why do you want 500 np inserts?

[Heisman]

because I think it will be good to detect variants and mutation. Also, I think that this much will be good enough to study SNP and gene expression.
What are your recommendations for some one who is doing WGS for the first time ?
what should one study if checking control vs treatment or normal vs diseased ?
How much should be insert and depth ?
May I know your opinion, if your guide gives you human samples (control vs treatment, and normal vs diseased) and if you are doing first time WGS ?
You may please write which may not be perfect but at least to start WGS with ?

I think you should start a new thread as I definitely won't be able to help you with RNA-seq stuff. I recommend you read a few reviews on the various experiments you are considering doing so you have a better idea of everything that goes into the data analysis and how to design the experiments.

[jp.]

Thank you very much for your help, however, may you recommend me something on WGS?

I think you should start a new thread as I definitely won't be able to help you with RNA-seq stuff. I recommend you read a few reviews on the various experiments you are considering doing so you have a better idea of everything that goes into the data analysis and how to design the experiments.

[jp.]

May you give me answer for the probblem posted here

http://seqanswers.com/forums/showthread.php?p=112098#post112098

Sorry I did not respond earlier. I am confused. Why do you want to build your own reference genome? What do you mean in your last post when you say you are going for 500 reads? You seem to want to do a lot of different experiments; do you have people around you who you can further discuss this with?

[LSerhGute]

When sequencing in the CLIA lab, are you doing proprietary panels used for diagnoses? or is it all still clinical research

[yongbao1]

we do clinical sequencing

[LSerhGute]

I realize my question may not have been clear. The reason for my question was that I was hoping to learn two things:

1.When doing clinical sequencing in the CLIA lab, are you doing targeted or WGS?

2. Was the sequencing test prescribed by a Dr. to diagnose a disease, and if so, did you lab develop the test?