Clustering plasma concentration-time curves: applications of unsupervised learning in pharmacogenomics.

Pharmaceutical researchers are continually searching for techniques to improve both drug development processes and patient outcomes. An area of recent interest is the potential for machine learning (ML) applications within pharmacology. One such application not yet given close study is the unsupervised clustering of plasma concentration-time curves, hereafter, pharmacokinetic (PK) curves. In this paper, we present our findings on how to cluster PK curves by their similarity. Specifically, we find clustering to be effective at identifying similar-shaped PK curves and informative for understanding patterns within each cluster of PK curves. Because PK curves are time series data objects, our approach utilizes the extensive body of research related to the clustering of time series data as a starting point. As such, we examine many dissimilarity measures between time series data objects to find those most suitable for PK curves. We identify Euclidean distance as generally most appropriate for clustering PK curves, and we further show that dynamic time warping, Fréchet, and structure-based measures of dissimilarity like correlation may produce unexpected results. As an illustration, we apply these methods in a case study with 250 PK curves used in a previous pharmacogenomic study. Our case study finds that an unsupervised ML clustering with Euclidean distance, without any subject genetic information, is able to independently validate the same conclusions as the reference pharmacogenomic results. To our knowledge, this is the first such demonstration. Further, the case study demonstrates how the clustering of PK curves may generate insights that could be difficult to perceive solely with population level summary statistics of PK metrics.

The impact of the Coronavirus pandemic on New York City real estate: First evidence.

We investigate whether pandemic-induced contagion disamenities and income effects arising due to COVID-related unemployment adversely affected real estate prices of one- or two-family owner-occupied properties across New York City (NYC). First, ordinary least squares hedonic results indicate that greater COVID case numbers are concentrated in neighborhoods with lower-valued properties. Second, we use a repeat-sales approach for the period 2003-2020, and we find that both the possibility of contagion and pandemic-induced income effects adversely impacted home sale prices. Estimates suggest sale prices fell by roughly $60,000 or around 8% in response to both of the following: 1000 additional infections per 100,000 residents and a 10-percentage point increase in unemployment in a given Modified Zip Code Tabulation Area (MODZCTA). These price effects were more pronounced during the second wave of infections. On the basis of cumulative MODZCTA infection rates through 2020, the estimated COVID-19 price discount ranged from approximately 1% to 50% in the most affected neighborhoods, and averaged 14%. The contagion effect intensified in the more affluent, but less densely populated NYC neighborhoods, while the income effect was more pronounced in the most densely populated neighborhoods with more rental properties and greater population shares of foreign-born residents. This disparity implies the pandemic may have been correlated with a wider gap in housing wealth in NYC between homeowners in lower-priced and higher-priced neighborhoods.