“We are witnessing something unique. With the wave of measures to control the spread of COVID-19, we have seen a wave of noise diminution. The world has become a quiet place.”
Thomas Lecocq made this striking observation at the fourth annual Ken Kennedy Institute Data Science Conference. The Oct. 26-28 event, held virtually, was hosted by the Ken Kennedy Institute and drew more than 300 leaders from industry and academia. The theme of the conference was shaped by COVID-19 and explored how the pandemic has influenced the use of data as well as the development and application of data science.
“Recognizing that discovery and innovation happens at the interfaces of disciplines and communities, the conference aims to bring together a diverse set of people from multiple communities spanning academia and industry,” said Lydia Kavraki, the Noah Harding Professor of Computer Science at Rice University and director of the Ken Kennedy Institute as she welcomed attendees to the conference.
“Data scientists will be future leaders and large contributors to development of new medicines and treatments,” said Reginald DesRoches, Rice provost and professor of civil engineering, “to using energy sources effectively, to creating safe and smart cities, and to helping understand and mitigate our changing climate.”
Lecocq, the conference’s opening speaker, is a geologist and seismologist with the Royal Observatory of Belgium who teaches at the Université Libre de Bruxelles. He collected data from 268 seismic stations in 117 countries and found significant noise reductions after the pandemic was declared in March by the World Health Organization, compared to before the global lockdown.
Human-generated noise customarily goes down during such quiet periods as Christmas/New Year week and the Chinese New Year, and during weekends and overnight. The drop in vibrations caused by the lockdown exceeds even those seen during these periods.
The cause: a lessening of human activity – motor vehicles, machinery, factories, even walking. The quiet permitted researchers to detect previously concealed earthquake signals, which helped them to distinguish human from natural seismic noise more clearly than ever before. Lecocq was the lead author of an article detailing his findings in the journal Science.
Lauren Stadler, an assistant professor of civil and environmental engineering at Rice, discussed her research in a talk titled “Spatial and Temporal Surveillance of Houston’s Wastewater to Track Community COVID-19 Infection Dynamics.”
The city of Houston treats some 250 million gallons of wastewater each day, serving more than 2.1 million people at 39 wastewater treatment plants. In collaboration with Houston Health Department, Houston Water, and Baylor College of Medicine, Stadler and her team collected and tested untreated samples from all the treatment plants.
SARS-CoV-2 was detected in samples taken from every plant. The data, Stadler said, suggests that detection of the virus in wastewater precedes positive tests in the population by one to two weeks in Houston. Stadler used the weekly data to identify geographic areas experiencing high infection rates and neighborhoods that may be experiencing an outbreak.
Houston Health Department COVID-19 teams use the wastewater viral trends identified by Stadler’s group to deploy testing units and contact tracers to areas where outbreaks were expected.
David Eagleman is director of the Center for Science and Law, adjunct professor at Stanford University, co-founder and CEO of Neosensory, Inc., and a 1993 graduate of Rice. In his talk, he discussed “How the Internet Will Save Civilization.”
“Why have previous civilizations folded up and collapsed?” he asked, citing 12 once powerful civilizations that no longer exist and suggesting these explanations: epidemics, natural disasters, poor information flow, political corruption, economic meltdown and resource depletion.
“We may be luckier than all of our predecessors, almost accidentally, because we have developed a rapid communications network that finds its highest expression in the internet. This technology obviates many of the threats faced by our ancestors. Our biggest risks may already be balanced by our most popular technology.”
The internet aided the response to COVID-19 in three ways: tele-presence, enabling people to work remotely and thus maintain social distancing; tele-medicine; and the optimal allocation of resources thanks to rapid tracking.
Dr. Nidal Moukaddam, associate professor at Baylor College of Medicine and director of the Ben Taub Hospital Psychiatry Outpatient Clinics, presented her talk “Help Me Feel Better: Leveraging Technology for Better Mental Health.” Moukaddam noted that the COVID-19 pandemic and subsequent lockdown have revealed “the fragility of good mental health and happiness in general.”
Higher rates of anxiety, depression and substance use have been observed since the start of the pandemic, and a rise in domestic violence reports. Her research, based on multiple studies of various populations, suggests that a “seamless integration of technology into mental health management requires operationalization of mental health measures.” Such measures are often subjective and perceived differently by patients and the people in their lives.
“Technology can create a direct interface with clinicians and adhere to healthcare-required confidentiality and privacy. Apps, wearables and basic internet connectivity can shape mental health and play an essential role in recognizing mental decline, diagnosis of mental illness and follow-up of symptomatic fluctuations in illness.
Her understanding is that in the first months of the pandemic, most people experienced some mental “deterioration,” but most of us have adapted to the new virus-imposed conditions.
“Technology can revolutionize how changes in behavior can help predict mental health issues and suggest how treatment should be delivered. But privacy and confidentiality issues abound and have to be addressed before we have successful partnerships,” Moukaddam said.
Also speaking were Ioannis Pavlidis, the Echkard-Pfeiffer Distinguished Professor of Computer Science at the University of Houston; Maria Chatzou Dunford, CEO and co-founder of Lifebit Biotech; Michelle Gill, senior scientist for deep learning and proteomics, and Avantika Lal, senior scientist for deep learning and genomics, both from NVIDIA; and Fritz Sedlazeck, assistant professor in the Human Genome Sequencing Center, Baylor College of Medicine. The conference featured a panel that discussed “Governing the Data Revolution” with Maria Chatzou Dunford, Lifebit Biotech; Caroline Chung, MD Anderson; David Jaffray, MD Anderson; Raghu Ramkrishnan, Microsoft; and Andre Dekker, MAASTRO Clinic, Maastricht University, Maastricht University Medical Center+. The panel exchanged views on the ways nations are developing privacy protection because while digital security technologies are rapidly maturing, technologies to record and govern data are far behind.
It also featured 13 technical talks on COVID-19, algorithms and foundations, business impact and healthcare. The chairs were David Jaffray, MD Anderson; Roy Keyes, Houston Data Science Group; Rodney Samaco, Baylor College of Medicine; Eric Venner, Human Genome Center, Baylor College of Medicine.
The conference included three-minute talks by 25 graduate students and postdocs who presented posters. The program was expanded to include a collaboration with MD Anderson, the Data-Driven Determinants for COVID-19 Oncology Discovery Effort (D3CODE). It was established to create a cross-functional, institution-wide data science initiative to understand cancer in context of the pandemic. The collaboration track included a panel with speakers from MD Anderson including David Jaffray, Caroline Chung, and Andy Futreal. They were joined by Rice faculty members, Lydia Kavraki and Chris Jermaine.
The conference committee included Denis Akhiyarov, Total; Natalie Berestovsky, Occidental Petroleum; Keith Cooper, Rice; Licong Cui, UTHealth; Scott Ferguson, HEDS Meetup; Luca Giancardo, UTHealth; Max Grossman, NAG; Giewee Hammond, Aramco Services; Xiaoqian Jiang, UTHealth; Zhandong Liu, Baylor College of Medicine; Scott Morton, Rice; Risa Myers, Rice; Jan E. Odegard, The Ion; Lilia Reddy, Chevron; Craig Rusin, Baylor College of Medicine; Akane Sano, Rice; Santiago Segarra, Rice; Rachel Shaffer, Chevron; Julianna Toms, BP; Tamas Toth, ExxonMobil; Jim Ward, Two Sigma; Yan Xu, Houston Machine Learning Meetup; Cheng Zhan, Microsoft; W. Jim Zheng, UTHealth.