Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
My startup: Genetic testing for CCR5 delta 32 (HIV Resistance) (hivgene.com)
40 points by kf on Jan 24, 2008 | hide | past | favorite | 53 comments


I think it's interesting and (if it were free) I'd want to know (even the me 10 years ago who wasn't married would).

But what problem does it solve? Why (besides curiosity) would I want that? Is it for HIV positive people to who might be hoping for a slow progression? For people who just want to feel a little more comfortable having lots of unprotected sex?

I guess I can see the former... Though I would think that the audience is relatively small (and hopefully getting smaller) and would be looking to their doctor for stuff like this.


For now, it is for information purposes only -- to satisfy curiosity. The state of the current research into CCR5 delta 32 doesn't allow for treatment decisions to be made -- it is known that CCR5 delta32 slows down the progression of HIV to AIDS, but you can't adjust medication dosage based on this. Similarly, scientists have concluded that individuals with CCR5 delta 32 in high risk populations are less likely to have HIV. But that study is based on an after the fact study of the group. It would be irresponsible for us to encourage unprotected sex and foolish for anyone to make a decision like that.

So what problem are we solving? I'm not sure that there is a real problem here, beyond a lack of genetic testing services. We're offering a service providing information and some people are more curious than others about this particular information.


Agreed. I know nothing about this gene, but I doubt it would be wise to go around having lots of unprotected sex just because a test from some random company said you were resistant to HIV. I think it would actually be irresponsible and dangerous if that were the explicit or implied message. A lawsuit waiting to happen. Make sure it's clear to your clients that that's not the case.


I could see you selling it to people who have been told they have AIDS, but want to see if they are actually immune and don't need the expensive (around 20,000 a year, last I've heard) medicine.


The market for that is incredibly small. I have only seen one case of a CCR5 delta 32 homozygous individual with HIV in the scientific literature. http://www.sciencedirect.com/science?_ob=ArticleURL&_udi...


If I read that article right it says 1% not a single case (but then again I didn't read too hard)


1% of people are homozygous for the mutation. The study itself is of one individual with the homozygous mutation that persistently had measurable viral loads for HIV-1. Normally the homozygous CCR5 mutation means you don't get HIV, because HIV infects through the CCR5 receptors which CCR5 D32 homozygous positive individuals don't have. But you could get infected by a blood transfusion from a late stage HIV patient, because HIV eventually starts infecting via the CXCR4 receptor. You could also get infected via the less transmitted and geographically confined HIV-2.

But these scenarios are unlikely to happen, so they haven't happened very often. The study was notable because they found ONE GUY who couldn't fight off HIV with CCR5 delta 32 homozyous.

If we could magically take away the CCR5 receptors from HIV+ individuals, many would be able to fight HIV to the point where they tested negative on some HIV tests.

See this old thread for discussion of Josh's curing HIV idea. http://news.ycombinator.com/item?id=47751


Thanks news.yc. What do you think?

This is our first site to be selling tests, we've got something of a business plan involving marketing different genetic tests to different audiences: men, women, b2b, etc. Send me an email if you'd like to see one of our other, flashier sites that isn't quite ready for launch.


* target audiences *

Craigslist m4m?


Despite your crudeness... that's actually not a bad idea...


I'm almost certain that that's what he meant (or something close to it). If not, then he hasn't looked into the relevant statistics, or doesn't understand marketing.

EDIT: Well, I guess he has looked in the relevant statistics.


Certainly men who have sex with men are a scientifically studied group with regards to CCR5 delta 32. We link to this study on the site. http://hsphsun3.harvard.edu/faculty/seage/JAIDS2001-27-p472....


Why is the term MSM used instead of "gay men"?


Because not all men who have sex with men consider themselves gay. And some gay men are celibate. For this study, at least, MSM is the most relevant term.


What does it matter if they don't consider themselves gay? If you smoke crack on a regular basis, you're a crackhead, regardless of whether you identify as one or not. Denial is irrelevant.

Some straight men are celibate. They still get called "heterosexual", though. Straight men who are not celibate are not called "men who have sex with women" (MSW?).


> What does it matter if they don't consider themselves gay?

They might be bisexual. The be trans-something. They might have other problems that make the sex and gender of their sex partners irrelevant: http://www1.villagevoice.com/people/0804,savage,78927,24.htm...

There is a good answer to your question having to do with the semiotics of sex and gender which is practical, smart, and important in day-to-day life (even if the people who wrote it all want to dig up the remains of Jacques Derrida so they can stare at his navel some more). I'm not going post it here because you probably don't want to read it, but you should know that the people who look down their noses at you for saying stuff like this aren't just doing it because of some feel-good PC nonsense.


>What does it matter if they don't consider themselves gay?

Well... this is mostly semantics here, but the term MSM is what makes the denial irrelevant.


> Some straight men are celibate. They still get called "heterosexual", though. Straight men who are not celibate are not called "men who have sex with women" (MSW?).

In fact, they are. Straight men who _are_ celibate would not be called MSW, and would not be at risk of contracting an STD. Did you bother to think this through before popping off about the use of what turns out to be the correct terminology?


Someone once told me you're only gay if you fall in love with a man. Having sex doesn't mean anything except you had sex with a man.


maybe they have sex with women, too. then "gay" doesn't fit so well.


They take away your Science Sounding Licence (SSL) if you fail to meet your TLA quota.


Granted this is not something most people probably would expect when they go for a checkup, but have you considered some kind of kit dentists could use or offer?


In the Castro I could see it becoming a huge hit... hell my old hairdresser could probably make a mint pushing these.


Is there any guarantee of confidentiality? I looked in the FAQ and saw nothing. This seems like a very sensitive test; I would not submit anything to someone who did not _absolutely guarantee_ confidentiality. Even then, I would be a little wary of doing something like this over the web.


Yes, privacy is of paramount importance. We would only disclose information if we were legally compelled to do so. I will add this to the FAQ.


You may even want to crank out some legalese about precisely what is done with the info: how long it is retained, what exact circumstances would compel you to release it, etc.

Sounds like you have the right idea, though.


Possible privacy hole:

What's to prevent someone from acquiring a sample of someone else's DNA and sending it off for you to test?


It's physically impractical to buccal swab someone without their consent.


Ah, that's more like it. :-)


1) What's to prevent someone else from selling a test for the same thing? Surely, you're not the first to consider a test for this mutation, right?

2) Also, I think you're going to have to be very careful with your copy. You're selling an unapproved genetic test, for which the only real purpose is to give people confidence that they can't be infected by HIV (don't split hairs -- this is how 99% of the population is going to view your service). The medical establishment is going to denounce you without a second thought, and I wouldn't be surprised if an enterprising lawyer tries to make a case against you for encouraging unsafe sex.

3) What are your error rates?

4) Your email input box is overlapping with your bullets under firefox 2 for mac.


1. There is little to nothing preventing competitors, though I would be surprised if anyone can compete with us on price.

2. I agree. A startup is inherently risky and we are taking every precaution we can.

3. Error rates are effectively 0% as far as the technique goes. The only possibility of error is human error in handling. Our testing procedures are specifically designed to minimize the risk of human error. Furthermore, due to the nature of the testing procedure, any possible errors would result in false negatives, not false positives. In addition, we have a policy of retesting all positive results.

4. Noted, thanks... it does the same thing in Safari also.


I'm suspicious of your claims in 3. What method are you using? Certainly not RFLP, right? Are you sequencing the gene? A labeled oligonucleotide, targeted toward the deletion?

I'm having a hard time conceiving of a method with zero error rate (either type I or II).


+1 It certainly beats that other junk you were trying to push:-)


You mean the memory stuff?...I forgot what it's called.


Leaving aside the doubtful utility of this test you really must provide some information about who is behind this company, what techniques will you use to perform the test? What expertise do you have in molecular diagnostics? How do potential customers know that you are not simply taking the money and sending out random results?


We've discussed putting how we do the test on the site, but concluded that we were best off not disclosing that information, seeing as no other genetic testing companies explain the techniques they use to do the testing. I guess you could call it a trade secret, even though anyone with experience in the field could figure it out pretty easily.

As for expertise in molecular diagnostics, I've been doing research at a hospital for a couple years now designing and developing tests such as this. Furthermore, most of these techniques are now taught in undergraduate level molecular biology labs.

For the last question, I suppose the same can be said for any company doing genetic testing. It would be unethical to simply take money and send out random results, though this could be a valid concern to some people. Do you have any suggestions on how we can prove the legitimacy of the results?


Disclosing who you are, qualifications and relevant experience would be good I think. If you don't feel comfortable putting them up then you'd have to understand customers might not feel comfortable using your service.


This is a great idea. I think that you should work on the presentation of the site, though. In the Web 2.0 world it's considered stylish or at least excusable to have a very barebones site, but the average person needs a more reassuring presence in order to trust someone with sensitive medical information.


This is something that we've discussed.

Actually, we've left the site looking "barebones" as an explicit design decision. People can be put-off by something "too slick," especially when you position yourself as a simple, single-product company, and especially in the research/academic world.

But you are right, too.

We are working on other brands with more premium product which we will release soon. The designs of these sites will reflect the more premium brand.

Email rms for more information if you're interested, he's our news.yc baron.


Barebones is nice, but it looks like a spam site or a template of some sort. It actually reduces customer confidence in that case.

Are you trying to follow the 23andme model, but only testing for a single gene?


Hey, I absolutely agree with you. I also think that McDonald's tastes like puke, but they run one hell of a franchise business.


Fix the "Huntingdon's disease" typo, for heaven's sake. You're asking people to view your company as an authority on the topic. Typos detract from that credibility.


You should use Authorize.net to process your payments...so your customer can stay on your site...I wouldnt like having to go to google to pay if I was going to buy.


I agree, it's on our to-do list. Though Google's $.20 + 2% - adwords discount is anticompetitively good.


+1 for 'anticompetitively good' :)


Doesn't google have a really nice API that you can use to make the process mostly transparent?


Old discussion on CCR5: http://news.ycombinator.com/item?id=47751

Includes Josh's idea for curing HIV


This site doesn't look trustworthy at all. Sorry..but for all I know you guys are just making up the results. Without more information on who you are, or how you do the testing, or some sort of certification from a trusted source, I wouldn't use this service because I don't know if the testing is accurate.


Is it possible to be diagnosed with HIV, not know you have CCR5, and mistakenly purchase the medicine?


Very, very unlikely. And you would still want to see a doctor and have your viral loads regularly checked.


Does 23andMe or anyone else in the space bundle testing for this gene in their larger/more-expensive packages?


No, 23andMe only tests for SNPs (Single Nucleotide Polymorphisms) and this is a 32 base pair deletion.


I forgot to include explicitly in my anti-23andMe rant that deletions, repeat polymorphisms, and the like were not included in their product's results. Thanks for reminding me :-)




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: