“In hindsight, the Human Genome Project stands alongside the Manhattan Project and the Apollo Program as one of the most successful mega-projects in the history of science.”

“It was a great time to be in genomics. I will say it was a lot of fun, it was a lot of stress.”

“Genomics really is a team science. It takes many perspectives. Find and listen to people who have different perspectives from yours.”

“If it saves the life of somebody you love, what matters most isn’t the details of the technology itself, but what the application of that technology can do for people.”

“I completely missed how big single cell genomics would be and how important it would be to science. That was my big surprise.”

“AI cannot solve all the problems. What we need to do is basically leverage the observed data plus the AI and also plus the statistics in order to accelerate the discovery.”

“So the goal is not to just have a discovery, but this had to be a true discovery.”

“It's not just about prediction. It's about intervention. What we should be thinking about is not just trying to detect cancer, but actually prevent it.”

“Cancer is not the only destination of somatic evolution. Some of this selection may actually protect against cancer.”

"Curing all disease within 5 or 10 years — I think that's just vastly overhyped. I think it's a bit counterproductive."

“If we have a robot that can walk into the lab, take out reagents from the freezer, unscrew Eppendorf tubes, and run a cell culture experiment successfully — then I think we're there.”

“Cancer is not one single disease, it's thousands of diseases.”

“Just like the periodic table is fundamental to chemistry, if you want to understand what's going on inside diseased tissue, you need a catalog of all the cell types within the tumor.”

“Despite being extremely high quality and impactful, the current human reference genome has key limitations—it is a fake genome representing a single mosaic haplotype. The human pangenome project is constructing phased, telomere-to-telomere diploid genome assemblies of actual individuals that represent broad human genetic diversity.”

“Human pan genome project, it's also a social science project. Ethical and legal issues are very, very important for us to consider.”

“DNA is the information source; the sequencer is the communication channel.”

“Through this journey I have enjoyed both the power of industry-driven engineering and the elegance of basic research.”

“We aim to make cost no longer a barrier for large-scale genomics through consumable reuse and chemical innovation.”

“Non-coding regions are not 'junk', but regulatory switches of disease.”

“We can not only fix genes, but also make cells 'starve' tumors by competing for nutrients.”

“3D structure is not a static scaffold, but a dynamic platform for transcriptional regulation.”

“Distant genes located within the nearby regions are not really transcribed independently; instead, they actually often can be clustered together in a proximity of spatial space, therefore providing the structural basis for coordinated transcription.”

“AI is underestimated. It's going to dramatically change everything. The underestimated part is creativity.”

“Inside every tumor lies the weapons to defeat it.”

“We use AI to harvest nature's engineering wonders. Instead of looking at a quintillion-size haystack, we just need to use AI to sort a curated short list.”

“The performance of cancer early screening technologies will undoubtedly be refined as data continues to accumulate. However, to advance population-level early cancer screening at the populational level, the health economics must be carefully calculated.”

“With innovative experimental high throughput technique developed at CPL, we are able to design Al to explore the understudied transcription factors.”

“The low quality data is also useful, but depending on the way how we use it.”

"From 0 to 1 is more about technology, clinical needs, and close collaboration with the clinical side. However, from 1 to 10 or from 10 to 100 is more about criteria and regulation. Each country has different regulations and environments, so technology must be localized to succeed."

“We need to treat regulation as a friend, not as someone that's trying to prevent us from doing things. Without regulation, no technology can go to scale.”