I (a geneticist) have to disagree. What you say is true for PCR as it is performed now, but a number of existing and rather routine quality control tests could be added to whatever law enforcement protocol is to circumvent this rather crude attempt to dupe the system.
You may be able to make a library, but you would have to do so in a YAC, BAC or plasmid (Yeast Artificial Chromosome, Bacterial Artificial Chromosome, and, well, plasmid). Such structures are detectable and discoverable.
Even if you performed mini-preps and purified out only the human segments your end sequences would be restriction enzyme cuts -- also discoverable via annealment and/or sequencing properties.
That and your end solution would be purified DNA in the absence of other human (or any!) DNA. You'd have to replicate the rest of the human genome sans the SRS sites that would otherwise be amplified to perform identification. Possible but not exactly trivial, especially if needing to do everything in the right relative quantities. Your planted sample would also need to have other genetic markers of your target â" example should the individual have blue eyes and brown hair, and you planted genomic DNA that called for brown eyes & hair, your sample would be exposed. Or say something more esoteric, like your target having a gene that predisposes them to a certain type of heart disease when your sample did not.
Next, you would need a human tissue sample -- blood, hair, semen, etc that has similarly been stripped of otherwise identifiable sequences. Absolutely non-trivial. Otherwise the lack of human cell-surface markers would be a pretty tell-tale sign that the sample was artificial, and if you used an unmodified source you would get a duplexed result.
Finally, single-cell PCR is now possible. In the lab we can isolate a single cell in a media bath and perform PCR on it. So in order to fool this sort of analysis you are going to have to perform transgenic modification on a cell culture to get your SRS library transfected into a living cell in the right quantity. Not only that but if with FISH analysis you would also have to get each SRS inserted into the correct chromosome and in the correct orientation, without leaving otherwise tell-tail signs of manipulation.
That would be a really tricky bit of work.
No, I think it would be a heck of a lot easier to try and obtain a living sample of the targetâ(TM)s cells, preferably something like a hair follicle, maybe get some epithelial cells from a discarded cup or some such. Trick would be to get that sample into a cell growth media quickly enough to rescue any still-living cells and then culture them. Of course then you have to again deal with purity issues â" a lot of the chemicals used in cell culturing would also be detectable even in trace amounts. Cell morphology for cells grown in culture tends to differ than that of naturally grown cells, etc. Overall this would be akin to planting a fingerprint from a discarded bit of trash, only a heck of a lot more complicated.
So yes, you could attempt to trick the system, and it may even work. But any number of assays that currently exist could be used to quickly identify any such attempt this sort of fraud.
Besides which, DNA evidence is currently classed as circumstantial evidence not direct, and cannot typically be used to convict on itâ(TM)s own merit. You still need to pair with the Big Three (means, motive, opportunity) as well as a preponderance of other bits of circumstantial or direct evidence.
