[Ovah]

If a DNA sample is taken from an assault victim, how is the DNA from the victim (technically) distinguished from that of perpetrator? Why isn't the readout an average of perpetrator and victim? Of course the perpetrator's readout should be saved but not the victim's.

At least where I am from (iirc) there is no sequencing involved. Instead the DNA is cut with restriction enzymes. Basically restriction enzyme + a persons DNA gives a unique fragment pattern.

[toomanyrichies]

Speculating here, but presumably the victim is willing and able to supply their own DNA as an elimination sample. Whatever DNA is left over gets run through a database, and possibly matched with an identity. If the identity comes with a photo, the victim may well be able to verify whether that photo is that of their attacker.

[kumarvvr]

DNA samples would have been taken from victim directly, for comparison. I guess samples and records also include location of sample collection (hair, skin, mouth, etc), time, date, circumstance, etc.

[aksss]

I assume that police departments will retain DNA from every sample they can get their hands on, just like fingerprints. You're in a DB at that point, and it's useful data (demonstrably). I'd be surprised if many police departments have a privacy policy, much less comply with it, regarding your rights after a DNA sample is taken, relative the auspices under which they collected it. Assume the worst if you're at all worried about your DNA being on file somewhere.

[Ovah]

I guess I could have been more clear. My question is about the lab side of things. Indefinite and indiscriminate data retention is definitely a problem although a different one.