Search

Differentiation of Crohn’s disease (CD) from intestinal tuberculosis (ITB) is a big challenge to gastroenterologists because of their indistinguishable features and insensitive diagnostic tools. A non-invasive biomarker is urgently required to distinguish ITB/CD patients particularly in India, a TB endemic region, where CD frequency is increasing rapidly due to urbanization. Among the three differentially expressed miRNAs obtained from small RNA transcriptomic profiling of ileocaecal/terminal ileal tissue of ITB/CD patients (n = 3), only two down-regulated miRNAs, miR-31-5p, and miR-215-5p showed comparable data in qRT-PCR. Out of which, only miR-215-5p was detectable in the patient’s plasma, but there was no significant difference in expression between ITB/CD. On the other hand, miR-375-3p, the pulmonary TB specific marker was found in higher amount in the plasma of ITB patients than CD while reverse expression was observed in the ileocaecal/terminal ileal tissues of the same patients. Next, using Bioplex pro-human cytokine 48-plex screening panel, only three chemokines, Eotaxin-1/CCL11, SDF-1α/CXCL12, and G-CSF have noted significantly different levels in the serum of ITB/CD patients. ROC analysis has revealed that compared to a single molecule, a combination of miR-375-3p + Eotaxin-1/CCL11 + SDF-1α /CXCL12 + G-CSF showed a better AUC of 0.83, 95% CI (0.69–0.96) with 100% specificity and positive predictive value while sensitivity, negative predictive value, and accuracy were 56%, 69%, and 78% respectively in distinguishing ITB from CD. This study suggests that a combination of plasma markers shows better potential in differentiating ITB from CD than a single marker and this panel of markers may be used for clinical management of ITB/CD patients.

Clustering ensemble, also referred to as consensus clustering, has emerged as a method of combining an ensemble of different clusterings to derive a final clustering that is of better quality and r...

Clustering has emerged as a method of unsupervised partitioning of a given set of data instances into a number of groups (called clusters) so that instances in the same group are more similar among each other with respect to instances in other groups. But there does not exist a universal clustering algorithm that can yield satisfactory result for any dataset. In this work we consider an ensemble (collection) of clusterings (partitions) of a dataset obtained in different ways and devise two methods that judiciously select clusters from different clusterings in the ensemble to construct a robust clustering. The superior performances of the proposed methods over well-known existing clustering algorithms on several benchmark datasets are empirically reported.

Consensus clustering, also called clustering ensemble, is a method of improving quality and robustness in clustering by optimally combining an ensemble of clusterings generated in different ways. In this work, we introduce our approach that is based on a selection-based model and use cumulative voting strategy in order to arrive at a consensus . We demonstrate the performance of our proposed method on several benchmark datasets and show empirically that it outperforms some well-known consensus clustering algorithms.

In this paper we propose a novel approach to document clustering by introducing a representative-based document similarity model that treats a document as an ordered sequence of words and partitions it into chunks for gaining valuable proximity information between words. Chunks are subsequences in a document that have low internal entropy and high boundary entropy. A chunk can be a phrase, a word or a part of word. We implement a linear time unsupervised algorithm that segments sequence of words into chunks. Chunks that occur frequently are considered as representatives of the document set. The representative based document similarity model, containing a term-document matrix with respect to the representatives, is a compact representation of the vector space model that improves quality of document clustering over traditional methods.