0:00:01.507,0:00:03.396 In 2003, 0:00:03.420,0:00:06.333 when we sequenced the human genome, 0:00:06.357,0:00:10.279 we thought we would have the answer[br]to treat many diseases. 0:00:10.974,0:00:13.631 But the reality is far from that, 0:00:14.782,0:00:16.703 because in addition to our genes, 0:00:16.727,0:00:21.297 our environment and lifestyle[br]could have a significant role 0:00:21.321,0:00:23.869 in developing many major diseases. 0:00:23.893,0:00:27.473 One example is fatty liver disease, 0:00:27.497,0:00:31.580 which is affecting over 20 percent[br]of the population globally, 0:00:31.604,0:00:34.638 and it has no treatment[br]and leads to liver cancer 0:00:34.662,0:00:36.085 or liver failure. 0:00:37.517,0:00:42.261 So sequencing DNA alone[br]doesn't give us enough information 0:00:42.285,0:00:44.517 to find effective therapeutics. 0:00:44.541,0:00:48.297 On the bright side, there are[br]many other molecules in our body. 0:00:48.321,0:00:52.301 In fact, there are[br]over 100,000 metabolites. 0:00:52.325,0:00:56.621 Metabolites are any molecule[br]that is supersmall in their size. 0:00:57.193,0:01:02.165 Known examples are glucose,[br]fructose, fats, cholesterol -- 0:01:02.189,0:01:03.699 things we hear all the time. 0:01:04.273,0:01:07.256 Metabolites are involved[br]in our metabolism. 0:01:08.066,0:01:12.094 They are also downstream of DNA, 0:01:12.118,0:01:17.200 so they carry information[br]from both our genes as well as lifestyle. 0:01:17.224,0:01:22.873 Understanding metabolites is essential[br]to find treatments for many diseases. 0:01:22.897,0:01:25.109 I've always wanted to treat patients. 0:01:25.934,0:01:29.792 Despite that, 15 years ago,[br]I left medical school, 0:01:29.816,0:01:31.781 as I missed mathematics. 0:01:33.019,0:01:35.955 Soon after, I found the coolest thing: 0:01:36.692,0:01:39.455 I can use mathematics to study medicine. 0:01:41.026,0:01:46.239 Since then, I've been developing[br]algorithms to analyze biological data. 0:01:47.092,0:01:49.375 So, it sounded easy: 0:01:49.399,0:01:53.000 let's collect data from all[br]the metabolites in our body, 0:01:53.024,0:01:58.152 develop mathematical models to describe[br]how they are changed in a disease 0:01:58.176,0:02:01.164 and intervene in those[br]changes to treat them. 0:02:02.488,0:02:05.960 Then I realized why no one[br]has done this before: 0:02:07.230,0:02:08.917 it's extremely difficult. 0:02:08.941,0:02:10.028 (Laughter) 0:02:10.052,0:02:12.464 There are many metabolites in our body. 0:02:12.783,0:02:15.283 Each one is different from the other one. 0:02:15.307,0:02:19.035 For some metabolites,[br]we can measure their molecular mass 0:02:19.059,0:02:21.652 using mass spectrometry instruments. 0:02:21.676,0:02:26.069 But because there could be, like,[br]10 molecules with the exact same mass, 0:02:26.093,0:02:27.900 we don't know exactly what they are, 0:02:27.924,0:02:30.698 and if you want to clearly[br]identify all of them, 0:02:30.722,0:02:33.826 you have to do more experiments,[br]which could take decades 0:02:33.850,0:02:35.564 and billions of dollars. 0:02:36.207,0:02:41.770 So we developed an artificial[br]intelligence, or AI, platform, to do that. 0:02:41.794,0:02:44.638 We leveraged the growth of biological data 0:02:44.662,0:02:49.086 and built a database of any existing[br]information about metabolites 0:02:49.110,0:02:52.238 and their interactions[br]with other molecules. 0:02:52.262,0:02:55.686 We combined all this data[br]as a meganetwork. 0:02:55.710,0:02:59.106 Then, from tissues or blood of patients, 0:02:59.130,0:03:01.881 we measure masses of metabolites 0:03:01.905,0:03:05.164 and find the masses[br]that are changed in a disease. 0:03:05.188,0:03:08.378 But, as I mentioned earlier,[br]we don't know exactly what they are. 0:03:08.402,0:03:13.537 A molecular mass of 180 could be[br]either the glucose, galactose or fructose. 0:03:13.561,0:03:15.580 They all have the exact same mass 0:03:15.604,0:03:17.691 but different functions in our body. 0:03:17.715,0:03:21.302 Our AI algorithm considered[br]all these ambiguities. 0:03:21.326,0:03:24.062 It then mined that meganetwork 0:03:24.086,0:03:28.439 to find how those metabolic masses[br]are connected to each other 0:03:28.463,0:03:30.421 that result in disease. 0:03:30.445,0:03:32.683 And because of the way they are connected, 0:03:32.707,0:03:37.030 then we are able to infer[br]what each metabolite mass is, 0:03:37.054,0:03:39.978 like that 180 could be glucose here, 0:03:40.002,0:03:42.553 and, more importantly, to discover 0:03:42.577,0:03:45.944 how changes in glucose[br]and other metabolites 0:03:45.968,0:03:47.473 lead to a disease. 0:03:47.497,0:03:50.492 This novel understanding[br]of disease mechanisms 0:03:50.516,0:03:55.008 then enable us to discover[br]effective therapeutics to target that. 0:03:55.601,0:03:59.446 So we formed a start-up company[br]to bring this technology to the market 0:03:59.470,0:04:01.275 and impact people's lives. 0:04:01.722,0:04:05.267 Now my team and I at ReviveMed[br]are working to discover 0:04:05.291,0:04:10.396 therapeutics for major diseases[br]that metabolites are key drivers for, 0:04:10.420,0:04:12.317 like fatty liver disease, 0:04:12.341,0:04:15.265 because it is caused[br]by accumulation of fats, 0:04:15.289,0:04:17.762 which are types[br]of metabolites in the liver. 0:04:17.786,0:04:21.726 As I mentioned earlier,[br]it's a huge epidemic with no treatment. 0:04:21.750,0:04:24.474 And fatty liver disease[br]is just one example. 0:04:24.498,0:04:28.676 Moving forward, we are going to tackle[br]hundreds of other diseases 0:04:28.700,0:04:30.193 with no treatment. 0:04:30.217,0:04:34.771 And by collecting more and more[br]data about metabolites 0:04:34.795,0:04:38.339 and understanding[br]how changes in metabolites 0:04:38.363,0:04:40.765 leads to developing diseases, 0:04:40.789,0:04:44.278 our algorithms will get[br]smarter and smarter 0:04:44.302,0:04:48.498 to discover the right therapeutics[br]for the right patients. 0:04:48.522,0:04:52.292 And we will get closer to reach our vision 0:04:52.316,0:04:56.179 of saving lives with every line of code. 0:04:56.203,0:04:57.524 Thank you. 0:04:57.548,0:05:01.375 (Applause)