Limits...
Composition and applications of focus libraries to phenotypic assays.

Wassermann AM, Camargo LM, Auld DS - Front Pharmacol (2014)

Bottom Line: In this article we discuss the types of compounds in these annotated libraries composed of tools, probes, and drugs.As well, we provide rationale and a few examples for how such libraries can enable phenotypic/forward chemical genomic approaches.As with any approach, there are several pitfalls that need to be considered and we also outline some strategies to avoid these.

View Article: PubMed Central - PubMed

Affiliation: Center for Proteomic Chemistry, Novartis Institutes for Biomedical Research Cambridge, MA, USA.

ABSTRACT
The wealth of bioactivity information now available on low-molecular weight compounds has enabled a paradigm shift in chemical biology and early phase drug discovery efforts. Traditionally chemical libraries have been most commonly employed in screening approaches where a bioassay is used to characterize a chemical library in a random search for active samples. However, robust curating of bioassay data, establishment of ontologies enabling mining of large chemical biology datasets, and a wealth of public chemical biology information has made possible the establishment of highly annotated compound collections. Such annotated chemical libraries can now be used to build a pathway/target hypothesis and have led to a new view where chemical libraries are used to characterize a bioassay. In this article we discuss the types of compounds in these annotated libraries composed of tools, probes, and drugs. As well, we provide rationale and a few examples for how such libraries can enable phenotypic/forward chemical genomic approaches. As with any approach, there are several pitfalls that need to be considered and we also outline some strategies to avoid these.

No MeSH data available.


Analysis of the entire bioactivity dataset. Examining biological annotations for the entire dataset including activity outside the extremes can rescue weak activity. (A) Depicted are three cases where compounds modulate cell signaling. In case 1, a compound shows high activity through, for example, activation of apoptosis. In case 2, three compounds are annotated to targets known to function in the same signaling pathway but each compound shows weak activity in the assay. In case 3, a compound shows strong activity in the assay acting through inhibition of general mechanisms such as transcription/translation. (B) While compounds in case 1 and 3 are easily captured using standard statistical thresholds, compounds in case 2 may only be rescued by examining the entire dataset and considering the target annotations in the context of biological pathway analysis. (C) Plot representing the aggregate effect of compounds perturbing a pathway to identify weakly active compounds (case 2). Each circle on the plot represents a pathway. The x-axis represents the aggregate effect of compounds hitting targets from the same pathway. For example, case 2 is depicted as having three compounds that individually show weak activity but are annotated as operating in the same pathway. The likelihood of getting such a score by chance, given the number of genes/proteins, is given on the y-axis (given as −log10, e.g., p = 0.01, 2). Using such an approach, relevant biological processes with subtle effects can be identified.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4109565&req=5

Figure 3: Analysis of the entire bioactivity dataset. Examining biological annotations for the entire dataset including activity outside the extremes can rescue weak activity. (A) Depicted are three cases where compounds modulate cell signaling. In case 1, a compound shows high activity through, for example, activation of apoptosis. In case 2, three compounds are annotated to targets known to function in the same signaling pathway but each compound shows weak activity in the assay. In case 3, a compound shows strong activity in the assay acting through inhibition of general mechanisms such as transcription/translation. (B) While compounds in case 1 and 3 are easily captured using standard statistical thresholds, compounds in case 2 may only be rescued by examining the entire dataset and considering the target annotations in the context of biological pathway analysis. (C) Plot representing the aggregate effect of compounds perturbing a pathway to identify weakly active compounds (case 2). Each circle on the plot represents a pathway. The x-axis represents the aggregate effect of compounds hitting targets from the same pathway. For example, case 2 is depicted as having three compounds that individually show weak activity but are annotated as operating in the same pathway. The likelihood of getting such a score by chance, given the number of genes/proteins, is given on the y-axis (given as −log10, e.g., p = 0.01, 2). Using such an approach, relevant biological processes with subtle effects can be identified.

Mentions: In addition to taking compound interference and chemical properties into consideration during the analysis of phenotypic screens, it is also important to understand how such focus library screens can be better interpreted. For example, unlike random screening, the expected hit rate in a focus set should be high as by design the compounds have been selected based on some prior knowledge (e.g., they target proteins in a pathway involved in the disease or phenotype of interest). Secondly, the definition of a “hit” is less clear as phenotypic screens measure a biological outcome that can result from the perturbation of different biological processes in the cell with varying degrees of potency. In biochemical target-based assay systems, high potency is a reasonable way to select actives. Conversely, when using cell-based/phenotypic assays, this approach, although still useful, can be problematic. This is because when screening a large collection of compounds that target different cellular processes, one may observe a large effect on the phenotype by compounds that simply target generic cellular mechanisms (e.g., cell-health related mechanisms, Figure 3) which, although these have genuine biological activity, are less interesting or non-specific. For example, it may be no surprise that HDAC inhibitors score in an assay that measures transcription (e.g., an RGA) or that proteasome inhibitors would have a strong effect on assays that measure protein production. Additionally, focusing on the strongest hits may result in missing compounds that effect specific nodes of a relevant pathway but have smaller effects in the assay. The smaller effect of such compounds can be attributed to several reasons: (1) position of the target on the pathway where compensatory mechanisms and pathway cross-talk may result in a minor perturbation of the phenotype; (2) lack of uniformity in the specificity and potency of compounds; (3) lack of uniform chemical properties across the library (cell permeability; binding kinetics), and (4) lack of congruence between the time required to observe a phenotypic effect by perturbing a given target vis-à-vis the duration of the screen. Therefore, the interpretation of phenotypic screens should take full advantage of prior biological knowledge, such as target and pathways, to facilitate analysis.


Composition and applications of focus libraries to phenotypic assays.

Wassermann AM, Camargo LM, Auld DS - Front Pharmacol (2014)

Analysis of the entire bioactivity dataset. Examining biological annotations for the entire dataset including activity outside the extremes can rescue weak activity. (A) Depicted are three cases where compounds modulate cell signaling. In case 1, a compound shows high activity through, for example, activation of apoptosis. In case 2, three compounds are annotated to targets known to function in the same signaling pathway but each compound shows weak activity in the assay. In case 3, a compound shows strong activity in the assay acting through inhibition of general mechanisms such as transcription/translation. (B) While compounds in case 1 and 3 are easily captured using standard statistical thresholds, compounds in case 2 may only be rescued by examining the entire dataset and considering the target annotations in the context of biological pathway analysis. (C) Plot representing the aggregate effect of compounds perturbing a pathway to identify weakly active compounds (case 2). Each circle on the plot represents a pathway. The x-axis represents the aggregate effect of compounds hitting targets from the same pathway. For example, case 2 is depicted as having three compounds that individually show weak activity but are annotated as operating in the same pathway. The likelihood of getting such a score by chance, given the number of genes/proteins, is given on the y-axis (given as −log10, e.g., p = 0.01, 2). Using such an approach, relevant biological processes with subtle effects can be identified.
© Copyright Policy - open-access
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4109565&req=5

Figure 3: Analysis of the entire bioactivity dataset. Examining biological annotations for the entire dataset including activity outside the extremes can rescue weak activity. (A) Depicted are three cases where compounds modulate cell signaling. In case 1, a compound shows high activity through, for example, activation of apoptosis. In case 2, three compounds are annotated to targets known to function in the same signaling pathway but each compound shows weak activity in the assay. In case 3, a compound shows strong activity in the assay acting through inhibition of general mechanisms such as transcription/translation. (B) While compounds in case 1 and 3 are easily captured using standard statistical thresholds, compounds in case 2 may only be rescued by examining the entire dataset and considering the target annotations in the context of biological pathway analysis. (C) Plot representing the aggregate effect of compounds perturbing a pathway to identify weakly active compounds (case 2). Each circle on the plot represents a pathway. The x-axis represents the aggregate effect of compounds hitting targets from the same pathway. For example, case 2 is depicted as having three compounds that individually show weak activity but are annotated as operating in the same pathway. The likelihood of getting such a score by chance, given the number of genes/proteins, is given on the y-axis (given as −log10, e.g., p = 0.01, 2). Using such an approach, relevant biological processes with subtle effects can be identified.
Mentions: In addition to taking compound interference and chemical properties into consideration during the analysis of phenotypic screens, it is also important to understand how such focus library screens can be better interpreted. For example, unlike random screening, the expected hit rate in a focus set should be high as by design the compounds have been selected based on some prior knowledge (e.g., they target proteins in a pathway involved in the disease or phenotype of interest). Secondly, the definition of a “hit” is less clear as phenotypic screens measure a biological outcome that can result from the perturbation of different biological processes in the cell with varying degrees of potency. In biochemical target-based assay systems, high potency is a reasonable way to select actives. Conversely, when using cell-based/phenotypic assays, this approach, although still useful, can be problematic. This is because when screening a large collection of compounds that target different cellular processes, one may observe a large effect on the phenotype by compounds that simply target generic cellular mechanisms (e.g., cell-health related mechanisms, Figure 3) which, although these have genuine biological activity, are less interesting or non-specific. For example, it may be no surprise that HDAC inhibitors score in an assay that measures transcription (e.g., an RGA) or that proteasome inhibitors would have a strong effect on assays that measure protein production. Additionally, focusing on the strongest hits may result in missing compounds that effect specific nodes of a relevant pathway but have smaller effects in the assay. The smaller effect of such compounds can be attributed to several reasons: (1) position of the target on the pathway where compensatory mechanisms and pathway cross-talk may result in a minor perturbation of the phenotype; (2) lack of uniformity in the specificity and potency of compounds; (3) lack of uniform chemical properties across the library (cell permeability; binding kinetics), and (4) lack of congruence between the time required to observe a phenotypic effect by perturbing a given target vis-à-vis the duration of the screen. Therefore, the interpretation of phenotypic screens should take full advantage of prior biological knowledge, such as target and pathways, to facilitate analysis.

Bottom Line: In this article we discuss the types of compounds in these annotated libraries composed of tools, probes, and drugs.As well, we provide rationale and a few examples for how such libraries can enable phenotypic/forward chemical genomic approaches.As with any approach, there are several pitfalls that need to be considered and we also outline some strategies to avoid these.

View Article: PubMed Central - PubMed

Affiliation: Center for Proteomic Chemistry, Novartis Institutes for Biomedical Research Cambridge, MA, USA.

ABSTRACT
The wealth of bioactivity information now available on low-molecular weight compounds has enabled a paradigm shift in chemical biology and early phase drug discovery efforts. Traditionally chemical libraries have been most commonly employed in screening approaches where a bioassay is used to characterize a chemical library in a random search for active samples. However, robust curating of bioassay data, establishment of ontologies enabling mining of large chemical biology datasets, and a wealth of public chemical biology information has made possible the establishment of highly annotated compound collections. Such annotated chemical libraries can now be used to build a pathway/target hypothesis and have led to a new view where chemical libraries are used to characterize a bioassay. In this article we discuss the types of compounds in these annotated libraries composed of tools, probes, and drugs. As well, we provide rationale and a few examples for how such libraries can enable phenotypic/forward chemical genomic approaches. As with any approach, there are several pitfalls that need to be considered and we also outline some strategies to avoid these.

No MeSH data available.