A major public health issue is the growing resistance of bacteria to antibiotics. An important part of the needed response is the discovery and development of new antimicrobial strategies. These require the screening of potential new drugs, typically accomplished using high-throughput screening (HTS). Traditionally, HTS is performed by examining one compound per well, but a more efficient strategy pools multiple compounds per well. In this work, we study several recently proposed pooling construction methods, as well as a variety of pooled high-throughput screening analysis methods, in order to provide guidance to practitioners on which methods to use. This is done in the context of an application of the methods to the search for new drugs to combat bacterial infection. We discuss both an extensive pilot study as well as a small screening campaign, and highlight both the successes and challenges of the pooling approach.
Pathogens are becoming increasingly resistant to antibiotics, and this fact has alarming implications for public health including increased mortality and associated healthcare costs [1,2]. The United Nations estimates that deaths due to antimicrobial resistance will be 10 million per year by 2050 [3]. For a specific example, the typhoidal strains of Salmonella enterica cause typhoid fever resulting in more than 100,000 deaths per year [4,5], and nontyphoidal strains are a leading cause of death from food-borne illness in the United States [6] and a contributor to the mortality in developing nations due to diarrhea [7,8]. Because of the surge in antibiotic-resistant typhoidal and non-typhoidal Salmonella enterica strains, the CDC and the WHO have called for the development of new drugs to address this need [1].
In response to this problem, a new anti-microbial strategy has been suggested [9] with potential in many bacteria including Salmonella enterica. An important part of developing that strategy is drug discovery, and the initial phase of this search is accomplished via high-throughput screening (HTS). High-throughput screening is used widely in drug discovery, chemical biology, and many other scientific and industrial domains. HTS involves the placement of specified compounds into wells in 384-or even 1536-well plates. The plates are then assayed and checked for a desired indication. Normally, a single compound is placed in each well, but various authors have suggested and shown that pooling multiple compounds in each well can improve throughput and/or statistical efficiency. Pooling has been controversial and contested, with both successes [e.g., 10,11,12,13] and cautions [14,15] in the literature. Recently, however, there have been reports of several successful pooled screening procedures [16,17,18]. Older methodologies, using approaches like orthogonal pooling [19] or poolHiTS [20], neither constructed their pools nor analyzed them using statistical methods, while the new approaches use statistical design ideas for pool construction [16,18] and statistical regularization for analysis [16,17,18].
In this work, we add to this emerging pooling literature in several ways. First, we describe in some detail a particular application related to a search for antimicrobials discussed at the outset (Section 2). In Section 3, we describe a number of existing pool construction methods and pooled HTS analysis methods, and make a set of extensive comparisons between them in Section 4. Most of the methods we consider are from the literature, but we propose a new Lasso thresholding method that exploits our knowledge of effect directions.
We also discuss a secondary analysis method to address a problem in these types of screens: they often produce too many false positives, which consume a large amount of resources.
The secondary criterion severely reduces the number of compounds to be validated, while still detecting large effects. In Section 5 we then present an extensive description of the pilot study (Section 5.1) used to establish pooling as a viable approach in this setting, as well as results from an initial screen (Section 5.2) which identified several promising compounds while reducing the number of considered false positives. We finish with a Discussion in Section 6.
Schwieters et al. [9] report that the enzyme mannitol-1-phosphate 5-dehydrogenase (MtlD) offers antimicrobial potential in many bacteria. This enzyme converts the compound mannitol-1-phosphate to fructose-6-phosphate. When mtlD mutant bacteria are provided mannitol, mannitol-1-phosphate accumulates, intoxicating the bacterium leading to reduced growth and attenuated virulence in animal models. Thus, the goal is to identify a compound that inhibits MtlD in the presence of mannitol. Using a wild-type bacterium, inhibition of MtlD in the presence of mannitol will result in lack of growth. However, numerous compounds will inhibit the growth of a wild-type bacterium for reasons unrelated to MtlD inhibition. To eliminate these from consideration, a parallel screen is used in which the bacterium cannot be harmed by a MtlD inhibitor (because it is a mtlA mutant and cannot form mannitol-1-phosphate). Thus, the screening problem is to find a drug which inhibits growth of the wild-type (WT) bacteria but not of the mtlA mutant (MUT) when assayed against both. We call such a drug a true hit. A drug which inhibits both WT and MUT is called a pseudo-hit. Our goal is to identify true hits.
Previous single-replicate, one-compound-one-well (OCOW) screening of 10,000 compounds for this system yielded 140 that appeared to inhibit WT, and 40 of those did not appear to inhibit MUT. However, upon retest in duplicate, none of the 140 were validated as true hits. Contemplating a similar screen scaled up to hundreds of thousands of compounds, the investigators realized that even a false positive rate of 1% would be extremely expensive. This led them to consider pooli
This content is AI-processed based on open access ArXiv data.