How many NGS readings for a virus genome? - (Nov/15/2015 )

We recently extracted RNA from our negative-sense virus and sent it to a company for NGS. The company called me today asking how many reads I need. I didn't know what to say :(

The virus genome is approximately 15 kb.

If it is only 15 kb, then you will likely get thousands of reads coverage with almost any number you give. But if they need a number, then typically people aim for 50x coverage. If the reads are 100 bp reads, then you need to cover 150 regions in a 15 kb genome, so you need around 7500 reads. It will be difficult to get so few. Tell them you need 50,000, which should be no problem.

Thanks so much. I will tell them I need 50,000.

They also asked me whether 1 gb of data is enough. After I explained what the sample really is he told me he will give me 2 gb. What did he mean by that? Did he mean giga bytes or giga bases? I was actually too shy to ask.  

This is the number of bases (which is similar in size to the number of bytes in a typical file, since 1 byte is used for 1 base). If you have 50,000 reads of 100 nt each, then you will have 5 gigabases total. So, number of reads * length of reads = total bases. I don't know the read length they are proposing. Common ones are 75 bp, 150 bp, 300 bp, and each of those with optional reverse reads. Any of those will work, but the easiest to assemble will be the longer bidirectional reads. I doubt you'll have assembly problems, although there could be  troublesome repeats. The forum to get good answers on this subject is Seqanswers.

They should be able to advise you next time, if you tell them what kind of assembly you need and what is the genome size (and what do you need it for, ie. do novo, resequencing, mutation detection..) and suggest the best read length and sample preparation.

At least we do advise your customers, since many of them don't have any prior experiences with NGS ;)

Thanks Trof. How long does it take to construct a library? It is now 3 weeks we've been waiting for our NGS result. Why does it take so long? We asked them to send the data to us, we will do the reading ourselves.

The library preparations, checkups and sequencing run can be done within week, week-and-half as I was able to observe. If you only want FASTQ files, they are produced immediately after run. But if it is a service company, they probably didn't just wait and do nothing, they may have schedule full for several weeks when your order comes. They should have informed you before ordering, when you can expect results.

But this is of course if everything goes fine.

We once waited in my old job over two months over results we needed for revision, but the library construction was suddenly failing and no one knew why. In the enzymatic preparation it sometimes happen for no obvious resons (though they are more quicker and less laborous the rest of the time). But if there is longer unexpected delay, they should inform you that they will not be able to make it in the time that was arranged.

Usually all commercial labs have wider time limits, so if there is a slight problem or delay, it won't affect the arranged time and they won't bother you. They just say it will be in a month, even when actually it takes a week. The same is for diagnostic lab, the actual time needed is less then stated, but there has to be some time margin in case they have full hands, something breaks or so. Also this may save some time for diagnostic tests or orders that are super urgent.