COMBImage2: a parallel computational framework for higher-order drug combination analysis that includes automated plate design, matched filter based object counting and temporal data mining

Abstract

Background: Pharmacological treatment of complex diseases using more than two drugs is commonplace in the clinic due to better efficacy, decreased toxicity and reduced risk for developing resistance. However, many of these higher-order treatments have not undergone any detailed preceding in vitro evaluation that could support their therapeutic potential and reveal disease related insights. Despite the increased medical need for discovery and development of higher-order drug combinations, very few reports from systematic large-scale studies along this direction exist. A major reason is lack of computational tools that enable automated design and analysis of exhaustive drug combination experiments, where all possible subsets among a panel of pre-selected drugs have to be evaluated.

Results: Motivated by this, we developed COMBImage2, a parallel computational framework for higher-order drug combination analysis. COMBImage2 goes far beyond its predecessor COMBImage in many different ways. In particular, it offers automated 384-well plate design, as well as quality control that involves resampling statistics and inter-plate analyses. Moreover, it is equipped with a generic matched filter based object counting method that is currently designed for apoptotic-like cells. Furthermore, apart from higher-order synergy analyses, COMBImage2 introduces a novel data mining approach for identifying interesting temporal response patterns and disentangling higher- from lower- and single-drug effects. COMBImage2 was employed in the context of a small pilot study focused on the CUSP9v4 protocol, which is currently used in the clinic for treatment of recurrent glioblastoma. For the first time, all 246 possible combinations of order 4 or lower of the 9 single drugs consisting the CUSP9v4 cocktail, were evaluated on an in vitro clonal culture of glioma initiating cells.

Conclusions: COMBImage2 is able to automatically design and robustly analyze exhaustive and in general higher-order drug combination experiments. Such a versatile video microscopy oriented framework is likely to enable, guide and accelerate systematic large-scale drug combination studies not only for cancer but also other diseases.

Keywords: Automated plate design; CUSP9v4; Data mining; Glioblastoma; Higher-order drug combination analysis; Label-free time-lapse video microscopy; MapReduce; Matched filter; Resampling.

Fig. 1

Experimental Capacity for Exhaustive Layouts. The experimental capacity required for performing exhaustive drug combination experiments grows rapidly with respect to the number of the individual drugs used. However, the graph shows that it is feasible to perform exhaustive experiments in one 384-well plate for up to 8 drugs

Fig. 2

COMBO-Pick flowchart. (1) A user-defined text specification file is imported; (2) Spatial feasibility control for 384-well format allowing at least 40 untreated wells is performed; (3) Alternative spatially feasible designs are suggested to the user; (4) Randomization of well destinations; (5) A plate destination specification for either exhaustive or pairwise drug combination experiments, compatible with Bridge, is produced per plate. 4-5 are repeated independently for all replicate plates, as specified by the user in (1)

Fig. 3

Randomized Plate Designs by COMBO-Pick. The pilot study was replicated 4 times using a differently randomized layout each time; R₁,R₂,R₃,R₄. Each layout consists of 5 different groups of wells based on the number of combined drugs: gray: 1; orange: 2; yellow: 3; cyan: 4; white: no drugs/untreated

Fig. 4

COMBO-MF Flowchart. THRESHOLD TUNING:(1)-(3) Matched filtering on training images; (4)-(5) Cross validation for optimal detection threshold. INTRA-PLATE ANALYSIS:(1) Image datastore selected by the user; (2) COMBO-Pick specification imported by the user; (3) MapReduce-based intra-plate quality control; (4)-(5) MapReduce-based quantification of apoptotic-like cells; (6) Table (CSV) with results; (7) Temporal graphics (EPS, PDF). INTER-PLATE ANALYSIS:(1) Intra-plate analysis employed separately for all replicates; (2) Results from (1) gathered and parsed; (3) Outlier removal based on the Inter-Plate QC as performed by COMBO-C; (4) Table (CSV) with merged inter-plate replicate values; (5) Temporal graphics (EPS, PDF)

Fig. 5

Apoptotic-like Object Counting. (a) Raw images where the prototypic object of size 33×32 pixels is overlaid on the left upper corner; (b) Prototypic-like detected objects. The green circles and orange crosses correspond to the detections made by the taboo- and position-based counting algorithms, respectively

Fig. 6

Prototypical Objects. COMBO-MF was evaluated using four similar prototypical objects of sizes: (a) 33×32 pixels; (b) 38×40 pixels; (c) 38×36 pixels and (d) 37×43 pixels

Fig. 7

COMBO-Mine Flowchart. (1) Results from all four analysis modules, COMBO-C, COMBO-M, COMBO-MF and COMBO-V, are required; (2) The extracted response patterns from (1) are organized into groups with similar behavior. For this grouping, multilevel K-means clustering is currently employed; (3) The smallest non-redundant subset of drugs and/or drug combinations for each group is identified by an exhaustive algorithmic search as shown in Fig. 8. (4) Each group is visualized by the corresponding average temporal profiles as determined by (2) and represented by the smallest non-redundant subset obtained from (3)

Fig. 8

Exhaustive Subset Search. Each group of the extracted response patterns is only represented by the smallest set of drugs and/or drug combinations that uniquely explains all of them in the same group. To illustrate the employed algorithmic procedure of this search, an example of our case study is used. In each iteration (iter 1−3), the drug combination of the lowest order is traced in all remaining ones of higher order. All higher-order combinations that include the to-be-traced lower-order combination, are subsequently removed. Thus, when the algorithm terminates, the corresponding non-redundant set of drug and drug combination names is formed

Fig. 9

COMBO-Mine Results. The employment of COMBO-Mine in the context of the current CUSP9 case study revealed 2 main groups with 2 subgroups each. Each (sub)group is visualized by the corresponding four average response patterns (three image based temporal profiles and one endpoint cell viability value) and characterized by the smallest non-redundant set of drugs and/or drug combinations in it

Fig. 10

Selected Microscopy Images for the CUSP9v4 case study. Example images of the sensitive GIC clone at t=68h after drug addition, based on the similarity grouping of the extracted response patterns as provided by COMBO-Mine. The images show: (1) untreated cells (shown twice), and cells treated with (2) Dis ; (3) (Min, Dis, Der, Que) ; (4) Apr ; (5) (Apr, Min)

Fig. 11

Inter-Plate Variability Estimate. The area spanned by the 4 growth curves is calculated and used as the inter-plate variability estimate for a particular well w. For convenience, this example illustrates the simplest case, where one replicate has the biggest change in cell growth for all time points