quantitative analysis it of colony morphology in yeast酵母菌落形态定量分析

quantitative analysis it of colony morphology in yeast酵母菌落形态定量 卢旺达饭店观后感 十七岁的单车观后感 quantitative analysis it of colony morphology in yeast酵母菌落形 态定量分析 Quantitative analysis of colony morphology in yeastPekka Ruusuvuori1,2, Jake Lin1,2,3, Adrian C. Scott4, Zhihao Tan4,6, Saija Sorsa1, AleksiKallio5, Matti Nykter5, Olli Yli-Harja1,2, Ilya Shmulevich1,2, and Aimée M. Dudley4,61Department of Signal Processing, Tampere University of Technology, Tampere, Finland2Institute for Systems Biology, Seattle, WA 3Luxembourg Centre for Systems Biomedicine,University of Luxembourg, Luxembourg 4Pacific Northwest Diabetes Research Institute, Seattle,WA 5Institute of Biomedical Technology, University of Tampere, Tampere, Finland 6Molecular andCellular Biology Program, University of Washington, Seattle, WAAbstractMicroorganisms often form multicellular structures such as biofilms and structured colonies thatcan influence the organism’s virulence, drug resistance, and adherence to medical devices.Phenotypic classification of these structures has traditionally relied on qualitative scoring systemsthat limit detailed phenotypic comparisons between strains. Automated imaging and quantitativeanalysis have the potential to improve the speed and accuracy of experiments designed to studythe genetic and molecular networks underlying different morphological traits. For this reason, wehave developed a platform that uses automated image analysis and pattern recognition to quantifyphenotypic signatures of yeast colonies. Our strategy enables quantitative analysis of individualcolonies, measured at a single time point or over a series of time-lapse images, as well as theclassification of distinct colony shapes based on image-derived features. Phenotypic changes incolony morphology can be expressed as changes in feature space trajectories over time, therebyenabling the visualization and quantitative analysis of morphological development. To facilitatedata exploration, results are plotted dynamically through an interactive Yeast Image Analysis webapplication (YIMAA; ) that integrates the raw and processed images across alltime points, allowing exploration of the image-based features and principal components associatedwith morphological development.Keyscolony morphology; image analysis; software; yeast; phenotype; time-lapseAddress correspondence to Aimée M. Dudley, Institute for Systems, Seattle, WA, aimee.dudley@gmail.com, or Pekka Ruusuvuori,Tampere University of Technology, Tampere, Finland, pekka.ruusuvuori@gmail.com.Supplementary material for this article is available at www.BioTechniques.com/article/114123.Author contributionsACS, ZT, and AMD designed the experiments. ACS and ZT performed the experiments. AMD supervised the experimental work. PR,JL, and IS designed software and computational analysis. JL, PR, SS, and AK wrote software. PR and JL performed the computationalanalysis. MN, OYH, and IS supervised the software development and computational work. PR, ACS, ZT, AMD, JL, MN, and ISwrote the paper. All authors read and approved the manuscript.Competing interestsThe authors declare no competing interests.NIH Public AccessAuthor ManuscriptBiotechniques. Author manuscript; available in PMC 2014 April 23.Published in final edited form as:Biotechniques. 2014 January ; 56(1): 18–27. doi:10.2144/000114123.NIH-PA Author ManuscriptNIH-PA Author ManuscriptNIH-PA Author Manuscript 天地玄黄观后感 有趣的汉字 夸父逐日扩写 好段摘抄大全 卢旺达饭店观后感 十七岁的单车观后感 A number of microorganisms, many of them well-known opportunistic pathogens, are ableto form highly structured biofilms and multicellular communities (1–4). The formation ofthese complex and well differentiated structures is thought to increase their resistance toantimicrobial treatments (5) and has been shown to be a key factor in persistent infections(1). Some strains of Saccharomyces cerevisiae, a non-pathogenic model organism, alsodisplay structured colony morphologies (5) with the characteristics of microbial biofilms,including the presence of an extracellular matrix composed largely of complexpolysaccharides (6–8), the development of channels in the colony interior (6), and the use ofcell-cell communication in colony development (9). The genetic tractability and numerous resources (10) not available for other biofilm forming organisms availabilityof makes S.cerevisiae an attractive organism in which to study the development of complexmorphologies, with the goal of ultimately uncovering the molecular mechanisms underlyingbiofilm formation (11).While studies aimed at characterizing the variation in colony morphology in S. cerevisiaehave been as objective as possible, qualitative classification schemes, such as having asingle investigator categorize colonies by eye, are still widely used (12–14). Image analysistools have also been applied to the automated analysis of yeast colonies. The image analysisplatform ImageJ (15) offers tools for processing and quantifying colony images (16), and theimage analysis tool CellProfiler (17) has been used to segment colonies on agar plates andgroup them based on shape, size, and color. Methods and software for quantifying colonygrowth combined with statistical analysis have also been presented in the literature (18,19).Other model organisms have also been subjected to quantitative, image-basedcharacterization and morphological classification. For example, image analysis has beenapplied to the automated screening of a variety of phenotypes (including morphology) inCaenorhabditis elegans (20), and recently an application similar to ours was applied to thestudy of filamentous fungi using a set of over 30 morphological features (21).Here, we describe an automated image analysis pipeline (Figure 1) that facilitates thequantitative study of colony morphology dynamics in large, time-lapse data sets. We startwith automated image processing and then extract a large, generic set of quantitativedescriptors. The combination of high-dimensional feature representation together with asparse, supervised logistic regression-based classification model is a powerful platform forthe analysis of colony morphology. We have also built a web-based application to facilitatethe intuitive exploration of the original raw and segmented time series images, the results ofPrincipal Component Analysis (PCA), and hundreds of individual quantitative features. Wetest the accuracy of our method by using it to computationally distinguish the complex(fluffy) and unstructured (smooth) colony phenotypes (6,22) based on image data from bothsingle time points and fine resolution time-lapses.Ruusuvuori et al.Page 2Biotechniques. Author manuscript; available in PMC 2014 April 23.NIH-PA Author ManuscriptNIH-PA Author ManuscriptNIH-PA Author Manuscript Materials and methodsYeast strains and growth conditionsStandard media and methods were used for the growth and genetic manipulation of S.cerevisiae (23). All colonies were grown and imaged in a 30?C warm room on YPD (2%glucose) agar plates. Strains used in this study are described in Table 1.Colony imagingColonies used to distinguish the fluffy and smooth phenotype based on a single time pointwere generated by manually micro-manipulating 天地玄黄观后感 有趣的汉字 夸父逐日扩写 好段摘抄大全 卢旺达饭店观后感 十七岁的单车观后感 individual cells into a gridded patternseparated by 10 mm in both the x- and y-axis. Colonies were imaged after five days ofgrowth using a PowerShot SX10IS camera outfitted with a Raynox DCR-250 macro lens(Yoshida Industry Co., Ltd. Tokyo, Japan).Colonies used for automated, time-lapse imaging were generated by depositing single cells12.7 mm apart in a checkerboard pattern with a FACSAria II cell sorter (BD Biosciences,Franklin Lakes, NJ) (Supplementary Materials). These colonies were imaged every 14 minfor 5 days using a 5d Mark II camera outfitted with a MP-E 65mm 1–5x macro lens(Cannon, Tokyo, Japan). The camera was attached to a custom built 2-axis gantry thatmoves the camera over the entire set of plates (Supplementary Materials). Camera settingswere held constant at an exposure time of 0.2 s and aperture of f/16. White balance was setusing a gray card. Focus was held constant.Generating quantitative colony phenotype signatures using image featuresThe first step in our automated pipeline involves segmenting the colony area as the region ofinterest (Supplementary Materials) and extracting features that describe the colony shape,size, intensity, fractal, and texture. We segment using a straightforward intensity-basedglobal thresholding operation (24) and then apply an additional size constraint to preventdetecting excessively small or large objects, which can arise from debris on the plate orcamera lens flare. We also perform image border clearing to remove false segmentationsthat occur when colonies located close to plate borders have refraction from the edge of theplate incorrectly assigned to the colony. This first set of segmentation masks (Figure 2A) isused for the first round of feature extraction. The shape and size categories include basicdescriptors for object morphology (e.g., area, convex area, and roundness). Intensity-basedfeatures provide measures of the intensity distribution (e.g., intensity percentilesand quantitative deviation), whereas the texture features [e.g., intensity deviations in local area, texturefeatures from gray-level co-occurrence matrices (25), histogram of oriented gradients (26),and local binary patterns (27)] take the spatial information into account.The next step involves an additional round of segmentation to detect shapes inside thecolonies, visible as intensity changes in 2-D projection images and the extraction of adifferent set of features from the segmented images. For this segmentation we use adifference of Gaussians segmentation (28), where the difference of two low-pass filteredversions of the original image (highly blurred and slightly blurred) is thresholded. The twolow-pass filters serve as a band-pass filter and the resulting binary image contains areasRuusuvuori et al.Page 3Biotechniques. Author manuscript; available in PMC 2014 April 23.NIH-PA Author ManuscriptNIH-PA Author ManuscriptNIH-PA Author Manuscript where intensity changes exist, but in which sharp variation, such as noise, is suppressed(Supplementary Materials). Ideally, the resulting segmentation mask would be empty for asmooth colony and capture the colony shape for a fluffy colony. The features extracted fromthese second segmentation masks include descriptors containing information about theshapes detected inside the colony (e.g., area of the mask relative to the colony size, maskarea in the center and border of the colony, number of objects in the mask, object sizes anddeviations).The combined feature set serves as a quantitative signature of colony phenotype, withcolonies derived from the same strain or belonging to the same phenotypic class sharingsimilar characteristics among many of the features (Figure 2D). A detailed description of all427 features is given in the Supplementary Materials. The feature list can be extended ortrimmed without changes to the subsequent classification 天地玄黄观后感 有趣的汉字 夸父逐日扩写 好段摘抄大全 卢旺达饭店观后感 十七岁的单车观后感 process.Supervised colony phenotype classificationTo transform these quantitative features into biologically meaningful phenotypeinformation, we used a supervised classification strategy. To circumvent the need to specifythe features used, we chose a classifier model with built-in feature selection, specifically thel1 regularized logistic regression (29,30), which produces sparse solutions and thus includesonly a subset of the features in the model.In logistic regression based classification, a feature vector x can be classified based on theconditional probability of belonging to the fluffy class given by the logistic regressionalgorithm as follows:[Eq. 1]where p(x) is the probability for the positive class given the feature vector x [i.e., p(x) =P(fluffy|X = x)], and the parameters β0 and β are estimated by maximizing the l1 penalizedlog-likelihood[Eq. 2]where F denotes the fluffy training samples, S is the smooth (non-fluffy) class trainingset, and λ is the class parameter regularizing the sparsity of the solution. In practice, the solutionis typically very sparse, leading to computationally efficient models (31), with only a smallsubset of features receiving a nonzero weight in vector β. Further, the use of logisticregression enables the extension to multi-class cases with more than two different strains orphenotypes.Quantitative analysis of colony spatiotemporal dynamicsTime-lapse image sequences are processed frame by frame as individual colony images oncethe colonies are large enough to be visible in the image (approximately one day of growth).The most obvious effect of colony growth is colony size, which also affects theRuusuvuori et al.Page 4Biotechniques. Author manuscript; available in PMC 2014 April 23.NIH-PA Author ManuscriptNIH-PA Author ManuscriptNIH-PA Author Manuscript quantification process. All features are extracted in the same manner from both small andlarge colonies. Feature trajectories are visualized by reducing the dimensionality withprincipal component analysis. Finally, a spatiotemporal profile of the yeast colony’ sdevelopment is built in which the spatial locations of the colony shapes are visualized overtime by taking a cumulative sum of the colony shape segmentation masks. Details can befound in the Supplementary Materials.Web application for data browsingWe have developed the Yeast Image Analysis (YIMAA) web application that serves as aninterface for the original and binary segmentation images together with the time-lapsedplotting of quantitative phenotypic results. YIMAA is built using the open sourcecomponents Highcharts. js, jQuery, and jQuery plugins. The design of YIMAA focuses oninteractivity and integration of images with dynamic time series plotting. Quantitativeresults are retrieved using AJAX. Image data are stored as assets organized by experimentand fetched on demand. The YIMAA web application is available at source code for the project, including the implementation of the image analysis pipelinecan be found at , we selected a general feature set that is nottailored to a single strain or classification task. Extracting a large set of image-derivedfeatures that measure different characteristics of the colony also helps ensure that changes inthe experiment or objects being studied (e.g., different magnifications, illumination settings,or strains) do not require significant alterations to the computational framework. Suchgeneralization will facilitate its use in a variety of applications.Our own research on yeast colony morphology has two experimental designs in which thisgeneral framework could be applied. First, the 天地玄黄观后感 有趣的汉字 夸父逐日扩写 好段摘抄大全 卢旺达饭店观后感 十七岁的单车观后感 classification of colonies into smooth andfluffy classes at a single time point, which was performed manually in our previous work(22), could be performed more objectively and in higher throughput using image-derivedfeatures. Second, an automated image analysis pipeline could be used to extract quantitativefeatures for many individual colonies as they grow and change shape over a series of time-lapsed images. In this framework, features extracted from the images form a vector ofnumerical values for each colony, where an element of the vector represents a feature valueat the time point sampled. Both descriptions of colony morphology could be used to informthe genetic analysis of a relatively large number of yeast strains under a variety ofenvironmental conditions.To assess the discriminating power of our morphological signatures, we first tested whetherthe method could distinguish the smooth and fluffy morphologies using static imagesacquired at a single time point (Figure 2). Smooth (YPG339, YPG 344, YPG348, YPG352,YPG356 and YPG360) and fluffy (F7, F11, F18, F25, F29, F31, F45, F47 and F49) yeaststrains (Table 1) were grown on solid YPD medium. Twenty replicates (colonies) of eachstrain were photographed daily, and day five was selected as the static time point. ColoniesRuusuvuori et al.Page 5Biotechniques. Author manuscript; available in PMC 2014 April 23.NIH-PA Author ManuscriptNIH-PA Author ManuscriptNIH-PA Author Manuscript that failed to grow were removed from subsequent analysis, yielding a data set of 251colony images. This data set was analyzed and uploaded to the YIMAA web application.Representative images are shown in Figure 2A, with a fluffy colony in the upper left and asmooth colony in the upper right. The ternary-valued segmentation images (below thecolony images) illustrate the region-of-interest identified by two rounds of segmentation,with the gray area corresponding to the intra-colony shapes (Methods). Quantitative featureswere then extracted from the images and normalized to zero-mean and unit variance.We determined the classification accuracy (98.79%) by performing a 4-fold crossvalidation for 5000 average repetitions with Monte Carlo random sampling on the 251 colonyimages described above. The upper panel of Figure 2B illustrates the distribution ofclassification accuracies for the validation partitions in the 5000 loop trials. The lower panelof Figure 2B shows the distribution of probability values (also obtained from the 5000 crossvalidation repetitions), where the probability of a sample x belonging to the fluffy class,p(x), is given by the logistic regression classifier. Classification is performed by dividing theprobability space into two classes. In practice, p(x) < 0.5 corresponds to a smoothclassification. Since the classifier is learned using 3/4 of the samples chosen randomly ateach repetition, the actual classification model varies between the trials and the values ofmodel weight vector β change within the validation loop. To analyze the model behaviorand learn which features are most informative, we collected the model parameter values inall 5000 trials. As expected, only a small number of features were used in the classifiermodel during the cross validation, with six features receiving a nonzero weight value in themodel weight vector β (Supplementary Materials).Next, we hierarchically clustered (in feature space) the colony image samples using thesubset of six features shown to contribute to the classifier model during cross validation. Theclustering (Figure 2C) showed a clear separation between the fluffy and smooth strains, andthe heat map reveals that colonies with the same phenotype share similar feature values. Theselection counts confirm that, as expected based on the applied regularization, the logisticregression classifier produced a sparse model using only a small subset of the features. Thus,the classification results obtained with the regularized 天地玄黄观后感 有趣的汉字 夸父逐日扩写 好段摘抄大全 卢旺达饭店观后感 十七岁的单车观后感 logistic regression classifier show thatthe features comprising phenotypic signatures can be used as a basis of classifying complexphenotypes in an automated manner when training samples are available.Interestingly, the histogram of probability values in Figure 2B appeared to consist of twomain distributions (large peaks on both the smooth and fluffy side) with additional, smallerpeaks on each side. Such behavior suggested the existence of phenotypic subclasses oroutlier samples. To explore this possibility, we analyzed the images that comprised thesesmall peaks manually and discovered that they corresponded to cases of respiratory deficientmutants (RDM) that had arisen spontaneously from the corresponding parental strain. Sincethe ability to respire drastically affects colony size as well as the ability to form fluffycolonies (22), we removed all images from RDM samples. Repeating the classificationprocedure described above on the remaining 238 images resulted in a near perfect averageclassification accuracy (Figure 2D), with only 5 false predictions out of 300,000classifications during cross validation. These probability distributions included only twomodes, and together with the improved classification accuracy, suggested that theRuusuvuori et al.Page 6Biotechniques. Author manuscript; available in PMC 2014 April 23.NIH-PA Author ManuscriptNIH-PA Author ManuscriptNIH-PA Author Manuscript respiratory deficient mutants were indeed not covered by the two-class model. Finally, wetested whether the logistic regression classification framework could be used to define athird class consisting of respiratory deficient mutants (13 samples). With a limited samplesize, we chose a simple leave-one-out cross validation, yielding 96.41% overall all fluffy and smooth samples classified correctly but only 4/13 RDM accuracy,with samplesclassified correctly. Thus, in this data set considering the RDM samples separately givesimproved classification accuracy for the fluffy and smooth phenotypes, but evaluating theapplicability of the proposed framework for automated classification of RDM sampleswould require a larger data set.To test the ability of the method to analyze the spatiotemporal dynamics of colonies as theygrow and change shape, we acquired a set of 18 time-lapse image sequences of 4 differentstrains (FY4, F29, F45 and YO779), where each sequence contained between 1 and 3colonies. Features were then extracted over the course of the time-lapse, providing aquantitative representation (in feature space) of the morphological dynamics of coloniesover time (Figure 3A). Examples of fluffy and smooth colonies at different times duringdevelopment are shown in Figure 3B. We also generated strain summaries for each strain ateach time point by taking the median value for each feature across all replicates. Both thefeature profiles of each individual replicate (colony) and these strain summaries were thenanalyzed by principal component analysis, allowing the trajectories in feature space as thecolony develops to be visualized in reduced dimensions (Figure 3C). The time-lapse results(Figure 3) demonstrate that the feature dynamics quantified for fluffy and smooth coloniesdiffer in the two example features, and the PCA plots reveal different feature trajectories fordifferent phenotype.In addition to the image analysis software, we also developed a web application (YIMAA,Supplementary Materials) that allows investigators to easily explore the results of thequantitative analysis alongside the raw input images from their experiment. The default pageplots the PCA analysis results for an example from this study (strain F29). Users can alsoselect multiple strains from the drop down list and their PCA results are plotted instantly.The plot can be animated to display points in order across the time series, allowing the userto explore the PCA values over time. This animation has 天地玄黄观后感 有趣的汉字 夸父逐日扩写 好段摘抄大全 卢旺达饭店观后感 十七岁的单车观后感 pause and play functions. As theplotting advances, the gallery container shows the raw and segmented image of the mostrecently plotted point. YIMAA can also plot a time series of any of the several hundredindividual features captured by the image analysis pipeline, and clicking on any time pointbrings up the associated images. Within the gallery panels, choosing a second strain permitsside-by-side image comparison. A user guide and screen shots of the YIMAA webapplication are included in the Supplementary Materials.Thus, we have developed a platform for the quantitative analysis of yeast colonymorphology and demonstrated its use for visualizing changes in colony morphology infeature space. We have also shown that these quantitative colony morphology signatures canbe used for supervised classification of colony phenotypes. These methods add statisticalrigor to the analysis of colony morphology and will enable the use of a variety ofcomputational tools, such as the classification and visualization tools described here, for theautomated analysis of colony shapes. The automated aspect of the software can also enableRuusuvuori et al.Page 7Biotechniques. Author manuscript; available in PMC 2014 April 23.NIH-PA Author ManuscriptNIH-PA Author ManuscriptNIH-PA Author Manuscript studies at scales not possible using manual scoring (i.e., extremely large numbers ofimages). Finally, a web application has been built for easy and rapid sharing of results. Thisintegrative environment for data exploration can be extended to other large-scale imageanalysis projects and to other colonyforming microorganisms.Supplementary MaterialRefer to Web version on PubMed Central for supplementary material.AcknowledgmentsThe authors thank Drs. Cecilia Garmendia-Torres and Alexander Skupin for helpful discussions, and TapioManninen for advice with data analysis. This work was funded by a National Institutes of Health Award (P50GM076547/Center for Systems Biology) and a strategic partnership between the Institute for Systems Biology andthe University of Luxembourg; P.R. is funded by Academy of Finland (project #140052); Z.T. is funded by theAgency for Science, Technology, and Research (Singapore). This paper is subject to the NIH Public Access Policy.References1. Costerton JW, Stewart PS, Greenberg EP. Bacterial biofilms: a common cause of persistentinfections. Science. 1999; 284:1318–1322. [PubMed: 10334980]2. Parsek MR, Greenberg EP. Sociomicrobiology: the connections between quorum sensing andbiofilms. Trends Microbiol. 2005; 13:27–33. [PubMed: 15639629]3. Lachke SA, Joly S, Daniels K, Soll DR. Phenotypic switching and filamentation in Candidaglabrata. Microbiology. 2002; 148:2661–2674. [PubMed: 12213913]4. Fries BC, Goldman DL, Cherniak R, Ju R, Casadevall A. Phenotypic switching in Cryptococcusneoformans results in changes in cellular morphology and glucuronoxylomannan structure. Infect.Immun. 1999; 67:6076–6083. [PubMed: 10531269]5. Cavalieri D, Townsend JP, Hartl DL. Manifold anomalies in gene expression in a vineyard isolate ofSaccharomyces cerevisiae revealed by DNA microarray analysis. Proc. Natl. Acad. Sci. USA. 2000;97:12369–12374. [PubMed: 11035792]6. Kuthan M, Devaux F, Janderova B, Slaninova I, Jacq C, Palkova Z. Domestication of wildSaccharomyces cerevisiae is accompanied by changes in gene expression and colony morphology.Mol. Microbiol. 2003; 47:745–754. [PubMed: 12535073]7. Karunanithi S, Vadaie N, Chavel CA, Birkaya B, Joshi J, Grell L, Cullen PJ. Shedding of the mucin-like flocculin Flo11p reveals a new aspect of fungal adhesion regulation. Curr. Biol. 2010; 20:1389–1395. [PubMed: 20619652]8. Váchová L, Stovicek V, Hlavacek O, Chernyavskiy O, Stepanek L, Kubinova L, Palkova Z. Flo11p,drug efflux pumps, and the extracellular matrix cooperate to form biofilm yeast colonies. J. 天地玄黄观后感 有趣的汉字 夸父逐日扩写 好段摘抄大全 卢旺达饭店观后感 十七岁的单车观后感 CellBiol. 2011; 194:679–687. [PubMed: 21875945]9. Vopálenská I, St’ovicek V, Janderova B, Vachova L, Palkova Z. Role of distinct dimorphictransitions in territory colonizing and formation of yeast colony architecture. Environ. Microbiol.2010; 12:264–277. [PubMed: 19799621]10. Botstein D, Fink GR. Yeast: an experimental organism for 21st Century biology. Genetics. 2011;189:695–704. [PubMed: 22084421]11. Reynolds TB, Fink GR. Bakers’ yeast, a model for fungal biofilm formation. Science. 2001;291:878–881. [PubMed: 11157168]12. Granek JA, Magwene PM. Environmental and genetic determinants of colony morphology inyeast. PLoS Genet. 2010; 6:e1000823. [PubMed: 20107600]13. St’oví?ek V, Vachova L, Kuthan M, Palkova Z. General factors important for the formation ofstructured biofilm-like yeast colonies. Fungal Genet. Biol. 2010; 47:1012–1022. [PubMed:20728557]Ruusuvuori et al.Page 8Biotechniques. Author manuscript; available in PMC 2014 April 23.NIH-PA Author ManuscriptNIH-PA Author ManuscriptNIH-PA Author Manuscript 14. Voordeckers K, De Maeyer D, van der Zande E, Vinces MD, Meert W, Cloots L, Ryan O, MarchalK, Verstrepen KJ. Identification of a complex genetic network underlying Saccharomycescerevisiae colony morphology. Mol. Microbiol. 2012; 86:225–239. [PubMed: 22882838]15. Schneider CA, Rasband WS, Eliceiri KW. NIH Image to ImageJ: 25 years of image analysis. Nat.Methods. 2012; 9:671–675. [PubMed: 22930834]16. Dymond JS, Richardson SM, Coombes CE, Babatz T, Muller H, Annaluru N, Blake WJ,Schwerzmann JW, et al. Synthetic chromosome arms function in yeast and generate phenotypicdiversity by design. Nature. 2011; 477:471–476. [PubMed: 21918511]17. Lamprecht MR, Sabatini DM, Carpenter AE. CellProfiler: free, versatile software for automatedbiological image analysis. Biotechniques. 2007; 42:71–75. [PubMed: 17269487]18. Memarian N, Jessulat M, Alirezaie J, Mir-Rashed N, Xu J, Zareie M, Smith M, Golshani A.Colony size measurement of the yeast gene deletion strains for functional genomics. BMCBioinformatics. 2007; 8:117. [PubMed: 17408490]19. Dittmar JC, Reid RJ, Rothstein R. Screen Mill: a freely available software suite for growthmeasurement, analysis and visualization of high-throughput screen data. BMC Bioinformatics.2010; 11:353. [PubMed: 20584323]20. W?hlby C, Kamentsky L, Liu ZH, Riklin-Raviv T, Conery AL, O’Rourke EJ, Sokolnicki KL,Visvikis O, et al. An image analysis toolbox for high-throughput C. elegans assays. Nat. Methods.2012; 9:714–716. [PubMed: 22522656]21. Posch AE, Spadiut O, Herwig C. A novel method for fast and statistically verified morphologicalcharacterization of filamentous fungi. Fungal Genet. Biol. 2012; 49:499–510. [PubMed:22587949]22. Tan Z, Hays M, Cromie GA, Jeffery EW, Scott AC, Ahyong V, Sirr A, Skupin A, Dudley AM.Aneuploidy underlies a multicellular phenotypic switch. Proc. Natl. Acad. Sci. USA. 2013;110:12367–12372. [PubMed: 23812752]23. Rose, MD.; Winston, FM.; Hieter, P. Cold Spring Harbor Laboratory. Methods in yeast genetics: alaboratory course manual. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press; 1990.24. Otsu N. A threshold selection method from gray-level histograms. IEEE Trans. Syst. Man Cybern.1979; 9:62–66.25. Haralick RM, Shanmugam K, Dinstein I. Textural Features for Image Classification. IEEE Trans.Syst. Man Cybern. 1973; SMC-3:610–621.26. Ludwig, O.; Delgado, D.; Gon?alves, V.; Nunes, U. Trainable classifier-fusion schemes: Anapplication to pedestrian detection; 12th International IEEE Conference on IntelligentTransportation Systems, ITSC ‘09; 2009. p. 1-6.27. Ojala T, Pietikainen M, Maenpaa T. Multiresolution gray-scale and rotation invariant textureclassification with local binary patterns. IEEE Trans. Pattern Anal. Mach. Intell. 2002; 24:971–987.28. Russ, JC. The image processing handbook. Boca Raton: CRC Press; 2011.29. 天地玄黄观后感 有趣的汉字 夸父逐日扩写 好段摘抄大全 卢旺达饭店观后感 十七岁的单车观后感 Tibshirani R. Regression shrinkage and selection via the lasso. J. R. Stat. Soc., B. 1996; 58:267–288.30. Friedman J, Hastie T, Tibshirani R. Regularization Paths for Generalized Linear Models viaCoordinate Descent. J. Stat. Softw. 2010; 33:1–22. [PubMed: 20808728]31. Manninen T, Huttunen H, Ruusuvuori P, Nykter M. Leukemia prediction using sparse logisticregression. PLoS ONE. 2013; 8:e72932. [PubMed: 24023658]32. Winston F, Dollard C, Ricupero-Hovasse SL. Construction of a set of convenient Saccharomycescerevisiae strains that are isogenic to S288C. Yeast. 1995; 11:53–55. [PubMed: 7762301]33. Fay JC, Benavides JA. Evidence for domesticated and wild populations of Saccharomycescerevisiae. PLoS Genet. 2005; 1:66–71. [PubMed: 16103919]34. Liti G, Barton DB, Louis EJ. Sequence diversity, reproductive isolation and species concepts inSaccharomyces. Genetics. 2006; 174:839–850. [PubMed: 16951060]Ruusuvuori et al.Page 9Biotechniques. Author manuscript; available in PMC 2014 April 23.NIH-PA Author ManuscriptNIH-PA Author ManuscriptNIH-PA Author Manuscript Method summaryOur platform enables the automated, quantitative analysis of yeast colony morphology byextracting a relatively large number of features from colony images followed bysupervised classification in feature space. This computational approach provides analternative to subjective scoring of colonies, is compatible with high-throughput -lapse experimental designs, and provides a web-based application for andtime dataexploration.Ruusuvuori et al.Page 10Biotechniques. Author manuscript; available in PMC 2014 April 23.NIH-PA Author ManuscriptNIH-PA Author ManuscriptNIH-PA Author Manuscript Figure 1. The components of the platform for automated quantitative analysis of yeast coloniesRuusuvuori et al.Page 11Biotechniques. Author manuscript; available in PMC 2014 April 23.NIH-PA Author ManuscriptNIH-PA Author ManuscriptNIH-PA Author Manuscript 2. Phenotype analysis of colonies from static images(A) Example images of fluffy and Figure smooth phenotypes and the corresponding segmentation results. (B) Classification accuracies(top) and probability values (bottom) for class representing the complex phenotypes during the 5000 repetitions. (C)Hierarchical clustering of the selected feature subspace shows how the features chosen by the logistic regression classifierseparate the phenotypes and how the colonies within... 天地玄黄观后感 有趣的汉字 夸父逐日扩写 好段摘抄大全