Machine Learning in Drug Discovery and Development Part 1: A Primer.

Artificial intelligence, in particular machine learning (ML), has emerged as a key promising pillar to overcome the high failure rate in drug development. Here, we present a primer on the ML algorithms most commonly used in drug discovery and development. We also list possible data sources, describe good practices for ML model development and validation, and share a reproducible example. A companion article will summarize applications of ML in drug discovery, drug development, and postapproval phase.

Challenges in Alzheimer's Disease Drug Discovery and Development: The Role of Modeling, Simulation, and Open Data.

Alzheimer's disease (AD) is the leading cause of dementia worldwide. With 35 million people over 60 years of age with dementia, there is an urgent need to develop new treatments for AD. To streamline this process, it is imperative to apply insights and learnings from past failures to future drug development programs. In the present work, we focus on how modeling and simulation tools can leverage open data to address drug development challenges in AD.

Growth rate analysis and efficient experimental design for tumor xenograft studies.

Human tumor xenograft studies are the primary means to evaluate the biological activity of anticancer agents in late-stage preclinical drug discovery. The variability in the growth rate of human tumors established in mice and the small sample sizes make rigorous statistical analysis critical. The most commonly used summary of antitumor activity for these studies is the T/C ratio. However, alternative methods based on growth rate modeling can be used. Here, we describe a summary metric called the rate-based T/C, derived by fitting each animal's tumor growth to a simple exponential model. The rate-based T/C uses all of the data, in contrast with the traditional T/C, which only uses a single measurement. We compare the rate-based T/C with the traditional T/C and assess their performance through a bootstrap analysis of 219 tumor xenograft studies. We find that the rate-based T/C requires fewer animals to achieve the same power as the traditional T/C. We also compare 14-day studies with 21-day studies and find that 14-day studies are more cost efficient. Finally, we perform a power analysis to determine an appropriate sample size.

IPM: An integrated protein model for false discovery rate estimation and identification in high-throughput proteomics.

In high-throughput mass spectrometry proteomics, peptides and proteins are not simply identified as present or not present in a sample, rather the identifications are associated with differing levels of confidence. The false discovery rate (FDR) has emerged as an accepted means for measuring the confidence associated with identifications. We have developed the Systematic Protein Investigative Research Environment (SPIRE) for the purpose of integrating the best available proteomics methods. Two successful approaches to estimating the FDR for MS protein identifications are the MAYU and our current SPIRE methods. We present here a method to combine these two approaches to estimating the FDR for MS protein identifications into an integrated protein model (IPM). We illustrate the high quality performance of this IPM approach through testing on two large publicly available proteomics datasets. MAYU and SPIRE show remarkable consistency in identifying proteins in these datasets. Still, IPM results in a more robust FDR estimation approach and additional identifications, particularly among low abundance proteins. IPM is now implemented as a part of the SPIRE system.

The United States of America and scientific research.

To gauge the current commitment to scientific research in the United States of America (US), we compared federal research funding (FRF) with the US gross domestic product (GDP) and industry research spending during the past six decades. In order to address the recent globalization of scientific research, we also focused on four key indicators of research activities: research and development (R&D) funding, total science and engineering doctoral degrees, patents, and scientific publications. We compared these indicators across three major population and economic regions: the US, the European Union (EU) and the People's Republic of China (China) over the past decade. We discovered a number of interesting trends with direct relevance for science policy. The level of US FRF has varied between 0.2% and 0.6% of the GDP during the last six decades. Since the 1960s, the US FRF contribution has fallen from twice that of industrial research funding to roughly equal. Also, in the last two decades, the portion of the US government R&D spending devoted to research has increased. Although well below the US and the EU in overall funding, the current growth rate for R&D funding in China greatly exceeds that of both. Finally, the EU currently produces more science and engineering doctoral graduates and scientific publications than the US in absolute terms, but not per capita. This study's aim is to facilitate a serious discussion of key questions by the research community and federal policy makers. In particular, our results raise two questions with respect to: a) the increasing globalization of science: "What role is the US playing now, and what role will it play in the future of international science?"; and b) the ability to produce beneficial innovations for society: "How will the US continue to foster its strengths?"

Estimating false discovery rates for peptide and protein identification using randomized databases.

MS-based proteomics characterizes protein contents of biological samples. The most common approach is to first match observed MS/MS peptide spectra against theoretical spectra from a protein sequence database and then to score these matches. The false discovery rate (FDR) can be estimated as a function of the score by searching together the protein sequence database and its randomized version and comparing the score distributions of the randomized versus nonrandomized matches. This work introduces a straightforward isotonic regression-based method to estimate the cumulative FDRs and local FDRs (LFDRs) of peptide identification. Our isotonic method not only performed as well as other methods used for comparison, but also has the advantages of being: (i) monotonic in the score, (ii) computationally simple, and (iii) not dependent on assumptions about score distributions. We demonstrate the flexibility of our approach by using it to estimate FDRs and LFDRs for protein identification using summaries of the peptide spectra scores. We reconfirmed that several of these methods were superior to a two-peptide rule. Finally, by estimating both the FDRs and LFDRs, we showed for both peptide and protein identification, moderate FDR values (5%) corresponded to large LFDR values (53 and 60%).

Some notable properties of the standard oncology Phase I design.

We identify three properties of the standard oncology Phase I trial design or 3 + 3 design. We show that the standard design implicitly uses isotonic regression to estimate a maximum tolerated dose. We next illustrate the relationship between the standard design and a Bayesian design proposed by Ji et al. (2007). A slight modification to this Bayesian design, under a particular model specification, would assign treatments in a manner identical to the standard design. We finally present calculations revealing the behavior of the standard design in a worst case scenario and compare its behavior with other 3 + 3-like designs.

Temporal global expression data reveal known and novel salicylate-impacted processes and regulators mediating powdery mildew growth and reproduction on Arabidopsis.

Salicylic acid (SA) is a critical mediator of plant innate immunity. It plays an important role in limiting the growth and reproduction of the virulent powdery mildew (PM) Golovinomyces orontii on Arabidopsis (Arabidopsis thaliana). To investigate this later phase of the PM interaction and the role played by SA, we performed replicated global expression profiling for wild-type and SA biosynthetic mutant isochorismate synthase1 (ics1) Arabidopsis from 0 to 7 d after infection. We found that ICS1-impacted genes constitute 3.8% of profiled genes, with known molecular markers of Arabidopsis defense ranked very highly by the multivariate empirical Bayes statistic (T(2) statistic). Functional analyses of T(2)-selected genes identified statistically significant PM-impacted processes, including photosynthesis, cell wall modification, and alkaloid metabolism, that are ICS1 independent. ICS1-impacted processes include redox, vacuolar transport/secretion, and signaling. Our data also support a role for ICS1 (SA) in iron and calcium homeostasis and identify components of SA cross talk with other phytohormones. Through our analysis, 39 novel PM-impacted transcriptional regulators were identified. Insertion mutants in one of these regulators, PUX2 (for plant ubiquitin regulatory X domain-containing protein 2), results in significantly reduced reproduction of the PM in a cell death-independent manner. Although little is known about PUX2, PUX1 acts as a negative regulator of Arabidopsis CDC48, an essential AAA-ATPase chaperone that mediates diverse cellular activities, including homotypic fusion of endoplasmic reticulum and Golgi membranes, endoplasmic reticulum-associated protein degradation, cell cycle progression, and apoptosis. Future work will elucidate the functional role of the novel regulator PUX2 in PM resistance.