Density-based Clustering Approach Research Articles

Density-Based Multiscale Analysis for Clustering in Strong Noise Settings With Varying Densities

Finding meaningful clustering patterns in data can be very challenging when the clusters are of arbitrary shapes, different sizes, or densities, and especially when the data set contains high percentage (e.g., 80%) of noise. Unfortunately, most existing clustering techniques cannot properly handle this tough situation and often result in dramatically deteriorating performance. In this paper, a purposefully designed clustering algorithm called Density-Based Multiscale Analysis for Clustering (DBMAC)-II is proposed, which is an improved version of the latest strong-noise clustering algorithm DBMAC. DBMAC is proposed under the assumption that all clusters are homogeneous and cannot work well when the data set contains clusters of varying densities. DBMAC-II overcomes the limitation of DBMAC by executing the multiscale analysis iteratively and can conduct strong noise-robust clustering without any strict assumption on the shapes and densities of clusters. In DBMAC-II, each data point or object is mapped into a feature space using its ${r}$ -neighborhood statistics with different ${r}$ (radius) values, which is similar to DBMAC. In general, the higher the value of ${r}$ -neighborhood statistics, the more likely the object is considered as a “clustered” object. Instead of trying to find a single optimal ${r}$ value, a set of radius values appropriate for separating “clustered” objects and “noisy” objects is identified, using a formal statistical method for multimodality test, referred to as multiscale analysis. For clusters with varying densities, multiscale analysis is applied to extract the clusters with the highest density from the current data set iteratively. Moreover, a statistical uniformity test for measuring clustering tendency is used as the self-adaptive stopping criterion of the iteration. Comprehensive experimental studies on a series of challenging benchmark data sets demonstrate that DBMAC-II is not only superior to classical density-based clustering approaches, including DBSCAN, OPTICS, and HDBSCAN, but also can consistently outperform the latest strong-noise robust clustering techniques, such as Skinny-dip.

Efficient parallel implementation of a density peaks clustering algorithm on graphics processing unit

The density peak (DP) algorithm has been widely used in scientific research due to its novel and effective peak density-based clustering approach. However, the DP algorithm uses each pair of data points several times when determining cluster centers, yielding high computational complexity. In this paper, we focus on accelerating the time-consuming density peaks algorithm with a graphics processing unit (GPU). We analyze the principle of the algorithm to locate its computational bottlenecks, and evaluate its potential for parallelism. In light of our analysis, we propose an efficient parallel DP algorithm targeting on a GPU architecture and implement this parallel method with compute unified device architecture (CUDA), called the ‘CUDA-DP platform’. Specifically, we use shared memory to improve data locality, which reduces the amount of global memory access. To exploit the coalescing accessing mechanism of GPU, we convert the data structure of the CUDA-DP program from array of structures to structure of arrays. In addition, we introduce a binary search-and-sampling method to avoid sorting a large array. The results of the experiment show that CUDA-DP can achieve a 45-fold acceleration when compared to the central processing unit based density peaks implementation.

An information-theoretic approach to hierarchical clustering of uncertain data

Uncertain data clustering has become central in mining data whose observed representation is naturally affected by imprecision, staling, or randomness that is implicit when storing this data from real-word sources. Most existing methods for uncertain data clustering follow a partitional or a density-based clustering approach, whereas little research has been devoted to the hierarchical clustering paradigm. In this work, we push forward research in hierarchical clustering of uncertain data by introducing a well-founded solution to the problem via an information-theoretic approach, following the initial idea described in our earlier work [26]. We propose a prototype-based agglomerative hierarchical clustering method, dubbed U-AHC, which employs a new uncertain linkage criterion for cluster merging. This criterion enables the comparison of (sets of) uncertain objects based on information-theoretic as well as expected-distance measures. To assess our proposal, we have conducted a comparative evaluation with state-of-the-art algorithms for clustering uncertain objects, on both benchmark and real datasets. We also compare with two basic definitions of agglomerative hierarchical clustering that are treated as baseline methods in terms of accuracy and efficiency of the clustering results, respectively. Main experimental findings reveal that U-AHC generally outperforms competing methods in accuracy and, from an efficiency viewpoint, is comparable to the fastest baseline version of agglomerative hierarchical clustering.

A novel DBSCAN with entropy and probability for mixed data

In big data situation, to detect clusters of different size and shape is a challenging and imperative task. Density based clustering approaches have been widely used in many areas of science due to its simplicity and the ability to detect clusters of different sizes and shapes over the last several years. With diverse conversion on categorical data, a modified version of the DBSCAN algorithm is proposed to cluster mixed data, noted as density based clustering algorithm for mixed data with integration of entropy and probability distribution (EPDCA). Optional and various conversions are provided for clustering process with adaptability. Some benchmark data sets from UCI have been selected for testing the capability and validity of EPDCA. It was shown that the clustering results of EPDCA are considerably improved, especially on automatically number of clusters formed, noise discovery and time elapsed to form clusters.

CC_TRS: Continuous Clustering of Trajectory Stream Data Based on Micro Cluster Life

The rapid spreading of positioning devices leads to the generation of massive spatiotemporal trajectories data. In some scenarios, spatiotemporal data are received in stream manner. Clustering of stream data is beneficial for different applications such as traffic management and weather forecasting. In this article, an algorithm for Continuous Clustering of Trajectory Stream Data Based on Micro Cluster Life is proposed. The algorithm consists of two phases. There is the online phase where temporal micro clusters are used to store summarized spatiotemporal information for each group of similar segments. The clustering task in online phase is based on temporal micro cluster lifetime instead of time window technique which divides stream data into time bins and clusters each bin separately. For offline phase, a density based clustering approach is used to generate macro clusters depending on temporal micro clusters. The evaluation of the proposed algorithm on real data sets shows the efficiency and the effectiveness of the proposed algorithm and proved it is efficient alternative to time window technique.

Phase and interface determination in computer simulations of liquid mixtures with high partial miscibility.

Partially miscible solutions can represent a challenge from the computer simulation standpoint, especially if the mutual solubility of the components is so large that their concentrations do not change much from one phase to another. In this case, identifying which molecules belong to which phase becomes a complicated task. Here, we propose a density-based clustering approach with self-tuning capabilities and apply it to the case of the mixture of an ionic liquid with benzene. The almost linear scaling of the algorithm makes it suitable for the analysis of long Molecular Dynamics or Monte Carlo trajectories.

Automated Clustering Analysis of Immunoglobulin Sequences in Chronic Lymphocytic Leukemia Based on 3D Structural Descriptors

Immunoglobulins (Igs) are crucial for the defense against pathogens, but they are also important in many clinical and biotechnological applications. Their characteristics, and ultimately their function, depend on their three-dimensional (3D) structure; however, the procedures to experimentally determine it are extremely laborious and demanding. Hence, the ability to gain insight into the structure of Igs at large relies on the availability of tools and algorithms for producing accurate Ig structural models based on their primary sequence alone. These models can then be used to determine structural and eventually functional similarities between different Igs. An example of such a task is the clustering of Igs based on their structure to determine meaningful common features such as the possible existence of common molecular targets (antigens). Several approaches have been proposed in order to achieve an optimal solution to this task yet their results were hindered mainly due to the lack of efficient clustering methods based on the similarity of 3D structure descriptors. Here, we present a novel workflow for robust Ig 3D modeling and automated clustering. We validated our protocol in chronic lymphocytic leukemia (CLL), where the clonotypic Igs are critically implicated in the disease ontogeny and evolution. Indeed, immunogenetic studies on the clonotypic Igs have strongly implicated antigen selection in the pathogenesis of CLL, while also providing robust prognostic information. In the present study, we used the structure prediction tools PIGS and I-TASSER for creating the 3D models and the TM-align algorithm to superpose them. The innovation of the current methodology resides in the usage of methods adapted from 3D content-based search methodologies to determine the local structural similarity between the 3D models. The Fast Point Feature Histograms descriptors derived from the structurally aligned parts are used to compute a distance matrix, which is then used as input for the clustering procedure. Clustering analysis on the data is performed through the application of the agglomerative and density-based clustering approaches. The first method is unsupervised whereas the second belongs to the semi-supervised type, i.e. requires a predefined number of clusters. To evaluate the quality of the herein described workflow, we performed a supervised analysis of 125 Ig 3D models originating from 5 CLL stereotyped subsets i.e. subgroups sharing (quasi) identical IGs, namely subsets #1, #2, #4, #6, #8. The reasoning behind this choice was that (i) homologous Ig primary sequences can be reasonably anticipated to be reflected in overall similar 3D structures, hence providing a reference for evaluating the developed workflow; and, (ii) these subsets are well characterized at both the clinical and biological levels. Subset size distribution was as follows: subset #1 (IGHV clan I/IGKV1(D)-39), n=37; subset #2 (IGHV3-21/IGLV3-21), n=43; subset #4 (IGHV4-34/IGKV2-30), n=22; subset #6 (IGHV1-69/IGKV3-20), n=12; and, subset #8 (IGHV4-39/IGKV1(D)-39), n=11. Overall, we obtained a high level of clustering accuracy i.e. Ig 3D model clusters matched to a very high degree the subsets defined by Ig primary sequence similarity. In detail, 5 Ig 3D model clusters were produced by: (i) cluster 1 containing 37/37 (100%) subset #1 models and one (8.3%) subset #6 model, (ii) cluster 2 containing 43/43 (100%) subset #2 models, (iii) cluster 3 containing 21/22 (95.5%) subset #4 models, (iv) cluster 4 containing 11/12 (91.7%) #6 models, and, (v) cluster 5 containing 11/11 (100%) subset #8 models along with a single (4.5%) subset #4 model (subsets #4 and #8 concern IgG CLL, in itself a rarity for CLL). These findings support that the innovative workflow described here enables robust clustering of 3D models produced from Ig sequences from patients with CLL. Furthermore, they indicate that CLL classification based on stereotypy of Ig primary sequences is likely also verified at the Ig 3D structural level. Studies are ongoing for both addressing the minor discrepancies observed here and producing the unsupervised 3D clustering of the IGs from a large series of both stereotyped and non-stereotyped CLL cases. DisclosuresRosenquist:Gilead Sciences: Speakers Bureau. Stamatopoulos:Gilead: Consultancy, Honoraria, Research Funding; Abbvie: Honoraria, Other: Travel expenses; Janssen: Honoraria, Other: Travel expenses, Research Funding; Novartis: Honoraria, Research Funding.

A Note on Fuzzy Joint Points Clustering Methods for Large Datasets

Integrating clustering algorithms with fuzzy logic typically yields more robust methods, which require little to no supervision of user. The fuzzy joint points method is a density-based fuzzy clustering approach that can achieve quality clustering. However, early versions of the method hold high computational complexity. In a recent work, the speed of the method was significantly improved without sacrificing clustering efficiency, and an even faster but parameter-dependent method was also suggested. Yet, the clustering performance of the latter was left as an open discussion and subject of study. In this study, we prove the existence of the appropriate parameter value and give an upper bound on it to discuss whether and how the parameter-dependent method can achieve the same clustering performance with the original method.

Fuzzy extensions of the DBScan clustering algorithm

The DBSCAN algorithm is a well-known density-based clustering approach particularly useful in spatial data mining for its ability to find objects' groups with heterogeneous shapes and homogeneous local density distributions in the feature space. Furthermore, it can be suitable as scaling down approach to deal with big data for its ability to remove noise. Nevertheless, it suffers for some limitations, mainly the inability to identify clusters with variable density distributions and partially overlapping borders, which is often a characteristics of both scientific data and real-world data. To this end, in this work, we propose three fuzzy extensions of the $$\textit{DBSCAN}$$DBSCAN algorithm to generate clusters with distinct fuzzy density characteristics. The original version of $$\textit{DBSCAN}$$DBSCAN requires two precise parameters (minPts and $$\epsilon $$∈) to define locally dense areas which serve as seeds of the clusters. Nevertheless, precise values of both parameters may be not appropriate in all regions of the dataset. In the proposed extensions of $$\textit{DBSCAN}$$DBSCAN, we define soft constraints to model approximate values of the input parameters. The first extension, named $$\textit{Fuzzy Core DBSCAN}$$Fuzzy CoreDBSCAN, relaxes the constraint on the neighbourhood's density to generate clusters with fuzzy core points, i.e. cores with distinct density; the second extension, named $$\textit{Fuzzy Border DBSCAN}$$Fuzzy BorderDBSCAN, relaxes $$\epsilon $$∈ to allow the generation of clusters with overlapping borders. Finally, the third extension, named $$\textit{Fuzzy DBSCAN}$$Fuzzy DBSCAN subsumes the previous ones, thus allowing to generate clusters with both fuzzy cores and fuzzy overlapping borders. Our proposals are compared w.r.t. state of the art fuzzy clustering methods over real-world datasets.

A Shared Nearest Neighbour Density based Clustering Approach on a Proclus Method to Cluster High Dimensional Data

Background/Objective: A high dimensional data is a dataset that ranges from a few to a hundreds of dimensions. Clustering such datasets needs an efficient algorithm such as Proclus but the algorithm has a drawback of ignoring cluster with small data points. So the proposed paper gives an ensemble of clustering that combines technique of two clustering algorithms to achieve a quality cluster of even small data points. Methods/Statistical Analysis: The research paper adapts a novel method of implementing a density based approach over a Proclus algorithm to cluster even small data points. These combined algorithms are tested using synthetic datasets. The Proclus algorithm is modified at a specific point where the density based algorithm is implemented. Findings: The results of the proposed algorithm are found to contain more clusters than mere Proclus algorithm does. The results is as such because in Proclus clustering the data point whose size is small are ignored so that only clusters with large number of data points can exists. However after the involvement of the shared nearest neighbor density based algorithm even the small data points are clustered which paves way for a more accurate and an efficient clustering process especially in a high dimensional data. Applications/Improvements: The application is a combination of two efficient algorithms but implemented in a simple way thereby reducing the complexity of the algorithm. The proposed technique can be applied on all high dimensional datasets irrespective of their sizes and shapes.

A density-based clustering approach for identifying overlapping protein complexes with functional preferences

BackgroundIdentifying protein complexes is an essential task for understanding the mechanisms of proteins in cells. Many computational approaches have thus been developed to identify protein complexes in protein-protein interaction (PPI) networks. Regarding the information that can be adopted by computational approaches to identify protein complexes, in addition to the graph topology of PPI network, the consideration of functional information of proteins has been becoming popular recently. Relevant approaches perform their tasks by relying on the idea that proteins in the same protein complex may be associated with similar functional information. However, we note from our previous researches that for most protein complexes their proteins are only similar in specific subsets of categories of functional information instead of the entire set. Hence, if the preference of each functional category can also be taken into account when identifying protein complexes, the accuracy will be improved.ResultsTo implement the idea, we first introduce a preference vector for each of proteins to quantitatively indicate the preference of each functional category when deciding the protein complex this protein belongs to. Integrating functional preferences of proteins and the graph topology of PPI network, we formulate the problem of identifying protein complexes into a constrained optimization problem, and we propose the approach DCAFP to address it. For performance evaluation, we have conducted extensive experiments with several PPI networks from the species of Saccharomyces cerevisiae and Human and also compared DCAFP with state-of-the-art approaches in the identification of protein complexes. The experimental results show that considering the integration of functional preferences and dense structures improved the performance of identifying protein complexes, as DCAFP outperformed the other approaches for most of PPI networks based on the assessments of independent measures of f-measure, Accuracy and Maximum Matching Rate. Furthermore, the function enrichment experiments indicated that DCAFP identified more protein complexes with functional significance when compared with approaches, such as PCIA, that also utilize the functional information.ConclusionsAccording to the promising performance of DCAFP, the integration of functional preferences and dense structures has made it possible to identify protein complexes more accurately and significantly.

Building Reconstruction From High-Resolution Multiview Aerial Imagery

In this letter, we propose a novel method to reconstruct accurate building structures from high-resolution multiview aerial imagery, using layered contour fitting (LCF) with a density-based clustering algorithm. Initially, the complicated 3-D scene is reconstructed by a probabilistic volumetric modeling algorithm. Subsequently, the reconstructed 3-D scene model is projected into layer images based on the height information. At last, we combine an extended layered density-based clustering approach with a generative LCF approach to remove noise and extract accurate building contours in every layer image at the same time. The final accurate 3-D building model is generated from these contours in layer images with a smoothing operation. Experiments on the aerial image sets demonstrate effectiveness and precision of our method.

Road-Network Aware Trajectory Clustering: Integrating Locality, Flow, and Density

Mining trajectory data has been gaining significant interest in recent years. However, existing approaches to trajectory clustering are mainly based on density and Euclidean distance measures. We argue that when the utility of spatial clustering of mobile object trajectories is targeted at road-network aware location-based applications, density and Euclidean distance are no longer the effective measures. This is because traffic flows in a road network and the flow-based density characterization become important factors for finding interesting trajectory clusters. We propose NEAT-a road-network aware approach for fast and effective clustering of trajectories of mobile objects traveling in road networks. Our approach carefully considers the traffic locality characterized by the physical constraints of the road network, the traffic flow among consecutive road segments, and the flow-based density to organize trajectories into spatial clusters in a comprehensive three-phase clustering framework. NEAT discovers spatial clusters as groups of sub-trajectories which describe both dense and highly continuous flows of mobile objects. We perform extensive experiments with mobility traces generated using different scales of real road networks. Experimental results demonstrate the flexibility of the NEAT system and show that NEAT is highly accurate and runs orders of magnitude faster than existing density-based trajectory clustering approaches.

Data Mining-based DNS Log Analysis

Domain name system (DNS) provides a critical function in directing Internet traffic. Defending DNS servers from bandwidth attacks is a significant task of DNS service providers. Traditional rule-based anomaly or intrusion detection methods are not able to update the rules dynamically. Data mining based approaches are able to find various patterns in the massive dynamic query traffic data. The patterns may assist the DNS service providers to detect anomalies in real time. In this paper, a novel frequent episode mining algorithm is proposed, as well as a volume trend prediction method which allows anomalies to be detected in real time. Density-based clustering approach is adopted to partition numerous domain names into different groups based on the characteristics of their query volume time series. Consistent episode mining method is proposed to find how the query traffic ‘propagate’ at different time between different domain names. Experiments are performed on a real-word DNS log data set. Interesting patterns are presented, indicating data mining based approaches are suitable and promising in the domain of DNS service.

Clustering Multi-Attribute Uncertain Data using Probability Distribution

Clustering is an unsupervised classification technique for grouping set of abstract objects into classes of similar objects. Clustering uncertain data is one of the essential tasks in mining uncertain data. Uncertain data is typically found in the area of sensor networks, weather data, customer rating data etc. The earlier methods for clustering uncertain data based on probability distribution, uses Kullback-Leibler divergence to measure similarity between two uncertain objects. In this paper, uncertain object in discrete domain is modeled, where uncertain object is treated as a discrete random variable. The JensonShannon divergence is used to measure the similarity between two uncertain objects and integrate it into partitioning and density based clustering approaches. Experiments are performed to verify the effectiveness and efficiency of model developed and results are at par with the existing approaches.

PiClust: A density based piRNA clustering algorithm

Piwi-interacting RNAs (piRNAs) are recently discovered, endogenous small non-coding RNAs. piRNAs protect the genome from invasive transposable elements (TE) and sustain integrity of the genome in germ cell lineages. Small RNA-sequencing data can be used to detect piRNA activations in a cell under a specific condition. However, identification of cell specific piRNA activations requires sophisticated computational methods. As of now, there is only one computational method, proTRAC, to locate activated piRNAs from the sequencing data. proTRAC detects piRNA clusters based on a probabilistic analysis with assumption of a uniform distribution. Unfortunately, we were not able to locate activated piRNAs from our proprietary sequencing data in chicken germ cells using proTRAC. With a careful investigation on data sets, we found that a uniform or any statistical distribution for detecting piRNA clusters may not be assumed. Furthermore, small RNA-seq data contains many different types of RNAs which was not carefully taken into account in previous studies. To improve piRNA cluster identification, we developed piClust that uses a density based clustering approach without assumption of any parametric distribution. In previous studies, it is known that piRNAs exhibit a strong tendency of forming piRNA clusters in syntenic regions of the genome. Thus, the density based clustering approach is effective and robust to the existence of non-piRNAs or noise in the data. In experiments with piRNA data from human, mouse, rat and chicken, piClust was able to detect piRNA clusters from total small RNA-seq data from germ cell lines, while proTRAC was not successful. piClust outperformed proTRAC in terms of sensitivity and running time (up to 200 folds). piClust is currently available as a web service at http://epigenomics.snu.ac.kr/piclustweb.

Leveraging microblogging big data with a modified density-based clustering approach for event awareness and topic ranking

Although diverse groups argue about the potential and true value benefits from social-media big data, there is no doubt that the era of big data exploitation has begun, driving the development of novel data-centric applications. Big data is notable not only because of its size, but also because of the complexity caused by its relationality to other data. In the past, owing to the limited possibilities of accessing big data, few data sources were available to allow researchers to develop advanced data-driven applications, such as monitoring of emerging real-world events. In fact, social media is greatly impacting the growth of big data; and big data is providing enterprises with the data to help them understand how to better detect marketing demands. Microblogging is a social network service capable of aggregating messages to explore facts and unknown knowledge. Nowadays, people often attempt to search for trending news and hot topics in real time from microblogging messages to satisfy their information needs. Under such a circumstance, a real demand is to find a way to allow users to organize a large number of microblogging messages into understandable events. In this work, we attempt to tackle such challenges by developing an online text-stream clustering approach using a modified density-based clustering model with collected microblogging big data. The system kernel combines three technical components, including a dynamic term weighting scheme, a neighbourhood generation algorithm and an online density-based clustering technique. After acquiring detected event topics by the system, our system provides functions for recommending top-priority event information to assist people to effectively organize emerging event data through the developed topic ranking algorithm.

Quality Prediction of Object Oriented Software Using Density Based Clustering Approach

A softw are fault prediction is a proven technique in achieving high software reliability. Prediction of fault-prone modules provides one way to support software quality engineering through improved scheduling and project control. Quality of software is increasingly important and testing related issues are becoming crucial for software. This necessitates the need to develop a real-time assessment technique that classifies these dynamically generated systems as being faulty/fault-free. A variety of software fault predictions techniques have been proposed, but none has proven to be consistently accurate. These techniques include statistical method, machine learning methods, parametric models and mixed algorithms. Therefore, there is a need to find the best techniques for Quality prediction of the software systems by finding the fault proneness. In this study, the performance of the Density-Based Spatial Clustering of Applications with Noise (DBSCAN) is evaluated for Java based Object Oriented Software system from NASA Metrics Data Program (MDP) data repository on the basis of fault proneness of the classes.

A density-based spatial clustering approach for defining local indicators of drinking water distribution pipe breakage

The physical condition of American infrastructure systems has raised concerns that have been addressed, in part, by studies addressing their condition assessment. Condition assessment aims at describing current condition and estimating remaining service lives of infrastructure network components. This is a predominantly time-based analysis, which can be complemented by the spatial analysis of physical condition data of infrastructure components. More specifically, exploratory spatial data analysis might identify areas with high failure rates and generate local indicators of condition for subsets of pipe segments within the physical network. Such local indicators can be used in cost/benefit analysis for planning capital investments with the advantage of allowing the identification of critical customers within critical regions and therefore better accounting for social costs. This paper aims to provide an approach to spatial data clustering of networked infrastructure failure data, which is presented and demonstrated by applying it to a drinking water pipe breakage dataset. Clusters are, in this paper, the set of break points that occurred in regions that present high breakage density per mile of pipe compared to the pipes in their vicinity. The proposed approach can be framed as density-based data clustering approach, and its output consists of a hierarchical clustering of breaks. The root node of the cluster hierarchy, which contains the set of all break points, is subsequently partitioned into smaller clusters. In this hierarchy, clusters are subdivided to reflect different breakage densities along the network space. Therefore, clusters in the lower level of the hierarchy present more homogeneous breakage rates. The results of the proposed approach are assessed according to their sensitivity to choice of parameters and according to a clustering quality measure. The chosen parameters are shown to provide results that are superior compared to a range of parameter choices, in terms of clustering quality, and also compared to the less important though relevant criteria of the number and size of clusters generated.

Elucidation of seventeen human peripheral blood B‐cell subsets and quantification of the tetanus response using a density‐based method for the automated identification of cell populations in multidimensional flow cytometry data

Advances in multiparameter flow cytometry (FCM) now allow for the independent detection of larger numbers of fluorochromes on individual cells, generating data with increasingly higher dimensionality. The increased complexity of these data has made it difficult to identify cell populations from high-dimensional FCM data using traditional manual gating strategies based on single-color or two-color displays. To address this challenge, we developed a novel program, FLOCK (FLOw Clustering without K), that uses a density-based clustering approach to algorithmically identify biologically relevant cell populations from multiple samples in an unbiased fashion, thereby eliminating operator-dependent variability. FLOCK was used to objectively identify seventeen distinct B-cell subsets in a human peripheral blood sample and to identify and quantify novel plasmablast subsets responding transiently to tetanus and other vaccinations in peripheral blood. FLOCK has been implemented in the publically available Immunology Database and Analysis Portal-ImmPort (http://www.immport.org)-for open use by the immunology research community. FLOCK is able to identify cell subsets in experiments that use multiparameter FCM through an objective, automated computational approach. The use of algorithms like FLOCK for FCM data analysis obviates the need for subjective and labor-intensive manual gating to identify and quantify cell subsets. Novel populations identified by these computational approaches can serve as hypotheses for further experimental study.