strains identification by STR micrsatellite using NGS

[fadss]

We have many vials that contain mixture of yeast STRAINS (of 3 or more strains). Currently we are using Capillaries Electrophoresis (CE). With culturing based method for isolation.
And to identify these strains we use microsatellite(STR) by amplifying 7-loci in multiplex PCR. So these are 7loci are non-overlapping and can be distinguished by different dye color in CE.
At the end of the day; we can find out how many&what strains are in each vial and in what proportion by comparing these microsatellite fingerprint to our database.

I am just wondering if there is a way we can use illumina or NGS to do this with much higher resolution and non-culturing method. Thus our current method current resolution are based on how many colonies we decided to isolate.
Our project are quite similar to forensic, but that we have mixture of different strains(individual if human).

Thanks!!

[SNPsaurus]

You can do this by NGS, but you really want to do it in a way that is cheaper and/or less laborious than now. Some sort of genotyping by sequencing (see http://arxiv.org/abs/1303.4835 for example) will let you multiplex hundreds of yeast samples in a HiSeq lane. If you have some reference strains as part of the set you could find SNPs that are unique to a particular strain, then go back to the mixed samples and look at those informative loci to see which strains are present.

To make it worthwhile you would want to make all the samples into libraries so you can do high multiplexing and bring down the cost of sequencing per samples. You could get by with a small # of loci (~1000), you'd want pretty high read depth at those loci since you have a mixed population (30-50X). That would let you multiplex 96-192 samples per lane safely.

Making the library, depending on the method, would be about the same as amplifying multiple loci.