Mohammed Alshalalfa

Efficient Periodicity Mining in Time Series Databases Using Suffix Trees

Periodic pattern mining or periodicity detection has a number of applications, such as prediction, forecasting, detection of unusual activities, etc. The problem is not trivial because the data to be analyzed are mostly noisy and different periodicity types (namely symbol, sequence, and segment) are to be investigated. Accordingly, we argue that there is a need for a comprehensive approach capable of analyzing the whole time series or in a subsection of it to effectively handle different types of noise (to a certain degree) and at the same time is able to detect different types of periodic patterns; combining these under one umbrella is by itself a challenge. In this paper, we present an algorithm which can detect symbol, sequence (partial), and segment (full cycle) periodicity in time series. The algorithm uses suffix tree as the underlying data structure; this allows us to design the algorithm such that its worstcase complexity is O(k.n2), where k is the maximum length of periodic pattern and n is the length of the analyzed portion (whole or subsection) of the time series. The algorithm is noise resilient; it has been successfully demonstrated to work with replacement, insertion, deletion, or a mixture of these types of noise. We have tested the proposed algorithm on both synthetic and real data from different domains, including protein sequences. The conducted comparative study demonstrate the applicability and effectiveness of the proposed algorithm; it is generally more time-efficient and noise-resilient than existing algorithms.

Characterization of 1577 primary prostate cancers reveals novel biological and clinicopathologic insights into molecular subtypes

BACKGROUND Prostate cancer (PCa) molecular subtypes have been defined by essentially mutually exclusive events, including ETS gene fusions (most commonly involving ERG) and SPINK1 overexpression. Clinical assessment may aid in disease stratification, complementing available prognostic tests. OBJECTIVE To determine the analytical validity and clinicopatholgic associations of microarray-based molecular subtyping. DESIGN, SETTING, AND PARTICIPANTS We analyzed Affymetrix GeneChip expression profiles for 1577 patients from eight radical prostatectomy cohorts, including 1351 cases assessed using the Decipher prognostic assay (GenomeDx Biosciences, San Diego, CA, USA) performed in a laboratory with Clinical Laboratory Improvements Amendment certification. A microarray-based (m-) random forest ERG classification model was trained and validated. Outlier expression analysis was used to predict other mutually exclusive non-ERG ETS gene rearrangements (ETS(+)) or SPINK1 overexpression (SPINK1(+)). OUTCOME MEASUREMENTS Associations with clinical features and outcomes by multivariate logistic regression analysis and receiver operating curves. RESULTS AND LIMITATIONS The m-ERG classifier showed 95% accuracy in an independent validation subset (155 samples). Across cohorts, 45% of PCas were classified as m-ERG(+), 9% as m-ETS(+), 8% as m-SPINK1(+), and 38% as triple negative (m-ERG(-)/m-ETS(-)/m-SPINK1(-)). Gene expression profiling supports three underlying molecularly defined groups: m-ERG(+), m-ETS(+), and m-SPINK1(+)/triple negative. On multivariate analysis, m-ERG(+) tumors were associated with lower preoperative serum prostate-specific antigen and Gleason scores, but greater extraprostatic extension (p<0.001). m-ETS(+) tumors were associated with seminal vesicle invasion (p=0.01), while m-SPINK1(+)/triple negative tumors had higher Gleason scores and were more frequent in Black/African American patients (p<0.001). Clinical outcomes were not significantly different among subtypes. CONCLUSIONS A clinically available prognostic test (Decipher) can also assess PCa molecular subtypes, obviating the need for additional testing. Clinicopathologic differences were found among subtypes based on global expression patterns. PATIENT SUMMARY Molecular subtyping of prostate cancer can be achieved using extra data generated from a clinical-grade, genome-wide expression-profiling prognostic assay (Decipher). Transcriptomic and clinical analysis support three distinct molecular subtypes: (1) m-ERG(+), (2) m-ETS(+), and (3) m-SPINK1(+)/triple negative (m-ERG(-)/m-ETS(-)/m-SPINK1(-)). Incorporation of subtyping into a clinically available assay may facilitate additional applications beyond routine prognosis.

Development and validation of a 24-gene predictor of response to postoperative radiotherapy in prostate cancer: a matched, retrospective analysis

BACKGROUND Postoperative radiotherapy has an important role in the treatment of prostate cancer, but personalised patient selection could improve outcomes and spare unnecessary toxicity. We aimed to develop and validate a gene expression signature to predict which patients would benefit most from postoperative radiotherapy. METHODS Patients were eligible for this matched, retrospective study if they were included in one of five published US studies (cohort, case-cohort, and case-control studies) of patients with prostate adenocarcinoma who had radical prostatectomy (with or without postoperative radiotherapy) and had gene expression analysis of the tumour, with long-term follow-up and complete clinicopathological data. Additional treatment after surgery was at the treating physicians discretion. In each cohort, patients who had postoperative radiotherapy were matched with patients who had not had radiotherapy using Gleason score, prostate-specific antigen concentration, surgical margin status, extracapsular extension, seminal vesicle invasion, lymph node invasion, and androgen deprivation therapy. We constructed a matched training cohort using patients from one study in which we developed a 24-gene Post-Operative Radiation Therapy Outcomes Score (PORTOS). We generated a pooled matched validation cohort using patients from the remaining four studies. The primary endpoint was the development of distant metastasis. FINDINGS In the training cohort (n=196), among patients with a high PORTOS (n=39), those who had radiotherapy had a lower incidence of distant metastasis than did patients who did not have radiotherapy, with a 10-year metastasis rate of 5% (95% CI 0-14) in patients who had radiotherapy (n=20) and 63% (34-80) in patients who did not have radiotherapy (n=19; hazard ratio [HR] 0·12 [95% CI 0·03-0·41], p<0·0001), whereas among patients with a low PORTOS (n=157), those who had postoperative radiotherapy (n=78) had a greater incidence of distant metastasis at 10 years than did their untreated counterparts (n=79; 57% [44-67] vs 31% [20-41]; HR 2·5 [1·6-4·1], p<0·0001), with a significant treatment interaction (pinteraction<0·0001). The finding that PORTOS could predict outcome due to radiotherapy treatment was confirmed in the validation cohort (n=330), which showed that patients who had radiotherapy had a lower incidence of distant metastasis compared with those who did not have radiotherapy, but only in the high PORTOS group (high PORTOS [n=82]: 4% [95% CI 0-10] in the radiotherapy group [n=57] vs 35% [95% CI 7-54] in the no radiotherapy group [n=25] had metastasis at 10 years; HR 0·15 [95% CI 0·04-0·60], p=0·0020; low PORTOS [n=248]: 32% [95% CI 19-43] in the radiotherapy group [n=108] vs 32% [95% CI 22-40] in the no radiotherapy group [n=140]; HR 0·92 [95% CI 0·56-1·51], p=0·76), with a significant interaction (pinteraction=0·016). The conventional prognostic tools Decipher, CAPRA-S, and microarray version of the cell cycle progression signature did not predict response to radiotherapy (pinteraction>0·05 for all). INTERPRETATION Patients with a high PORTOS who had postoperative radiotherapy were less likely to have metastasis at 10 years than those who did not have radiotherapy, suggesting that treatment with postoperative radiotherapy should be considered in this subgroup. PORTOS should be investigated further in additional independent cohorts. FUNDING None.

Prediction of novel drug indications using network driven biological data prioritization and integration.

BackgroundWith the rapid development of high-throughput genomic technologies and the accumulation of genome-wide datasets for gene expression profiling and biological networks, the impact of diseases and drugs on gene expression can be comprehensively characterized. Drug repositioning offers the possibility of reduced risks in the drug discovery process, thus it is an essential step in drug development.ResultsComputational prediction of drug-disease interactions using gene expression profiling datasets and biological networks is a new direction in drug repositioning that has gained increasing interest. We developed a computational framework to build disease-drug networks using drug- and disease-specific subnetworks. The framework incorporates protein networks to refine drug and disease associated genes and prioritize genes in disease and drug specific networks. For each drug and disease we built multiple networks using gene expression profiling and text mining. Finally a logistic regression model was used to build functional associations between drugs and diseases.ConclusionsWe found that representing drugs and diseases by genes with high centrality degree in gene networks is the most promising representation of drug or disease subnetworks.

Discovery and Validation of Novel Expression Signature for Postcystectomy Recurrence in High-Risk Bladder Cancer

Background Nearly half of muscle-invasive bladder cancer patients succumb to their disease following cystectomy. Selecting candidates for adjuvant therapy is currently based on clinical parameters with limited predictive power. This study aimed to develop and validate genomic-based signatures that can better identify patients at risk for recurrence than clinical models alone. Methods Transcriptome-wide expression profiles were generated using 1.4 million feature-arrays on archival tumors from 225 patients who underwent radical cystectomy and had muscle-invasive and/or node-positive bladder cancer. Genomic (GC) and clinical (CC) classifiers for predicting recurrence were developed on a discovery set (n = 133). Performances of GC, CC, an independent clinical nomogram (IBCNC), and genomic-clinicopathologic classifiers (G-CC, G-IBCNC) were assessed in the discovery and independent validation (n = 66) sets. GC was further validated on four external datasets (n = 341). Discrimination and prognostic abilities of classifiers were compared using area under receiver-operating characteristic curves (AUCs). All statistical tests were two-sided. Results A 15-feature GC was developed on the discovery set with area under curve (AUC) of 0.77 in the validation set. This was higher than individual clinical variables, IBCNC (AUC = 0.73), and comparable to CC (AUC = 0.78). Performance was improved upon combining GC with clinical nomograms (G-IBCNC, AUC = 0.82; G-CC, AUC = 0.86). G-CC high-risk patients had elevated recurrence probabilities (P < .001), with GC being the best predictor by multivariable analysis (P = .005). Genomic-clinicopathologic classifiers outperformed clinical nomograms by decision curve and reclassification analyses. GC performed the best in validation compared with seven prior signatures. GC markers remained prognostic across four independent datasets. Conclusions The validated genomic-based classifiers outperform clinical models for predicting postcystectomy bladder cancer recurrence. This may be used to better identify patients who need more aggressive management.

Racial Variations in Prostate Cancer Molecular Subtypes and Androgen Receptor Signaling Reflect Anatomic Tumor Location.

UNLABELLED Prostate cancer (PCa) subtypes based on ETS gene expression have been described. Recent studies suggest there are racial differences in tumor location, with PCa located anteriorly more often among African-American (AA) compared to Caucasian-American (CA) men. In this retrospective analysis of a multi-institutional cohort treated by radical prostatectomy (179 CA, 121 AA), we evaluated associations among molecular subtype, race, anatomic tumor location, and androgen receptor (AR) signaling. Subtype (m-ERG(+), m-ETS(+), m-SPINK1(+), or triple-negative) was determined using distribution-based outlier analysis. AR signaling was investigated using gene expression profiling of canonical AR targets. m-ERG(+) was more common in CA than AA men (47% vs 22%, p<0.001). AA men were more likely to be m-SPINK1(+) (13% vs 7%; p=0.069) and triple-negative (50% vs 37%; p=0.043). Racial differences in molecular subtypes did not persist when tumors were analyzed by location, suggesting a biologically important relationship between tumor location and subtype. Accordingly, anterior tumor location was associated with higher Decipher scores and lower global AR signaling. PATIENT SUMMARY This study demonstrates associations among patient race, prostate cancer molecular subtypes, and tumor location. Location-specific differences in androgen regulation may further underlie these relationships.

Association of multiparametric MRI quantitative imaging features with prostate cancer gene expression in MRI-targeted prostate biopsies

Standard clinicopathological variables are inadequate for optimal management of prostate cancer patients. While genomic classifiers have improved patient risk classification, the multifocality and heterogeneity of prostate cancer can confound pre-treatment assessment. The objective was to investigate the association of multiparametric (mp)MRI quantitative features with prostate cancer risk gene expression profiles in mpMRI-guided biopsies tissues. Global gene expression profiles were generated from 17 mpMRI-directed diagnostic prostate biopsies using an Affimetrix platform. Spatially distinct imaging areas (‘habitats’) were identified on MRI/3D-Ultrasound fusion. Radiomic features were extracted from biopsy regions and normal appearing tissues. We correlated 49 radiomic features with three clinically available gene signatures associated with adverse outcome. The signatures contain genes that are over-expressed in aggressive prostate cancers and genes that are under-expressed in aggressive prostate cancers. There were significant correlations between these genes and quantitative imaging features, indicating the presence of prostate cancer prognostic signal in the radiomic features. Strong associations were also found between the radiomic features and significantly expressed genes. Gene ontology analysis identified specific radiomic features associated with immune/inflammatory response, metabolism, cell and biological adhesion. To our knowledge, this is the first study to correlate radiogenomic parameters with prostate cancer in men with MRI-guided biopsy.

Integrated classification of prostate cancer reveals a novel luminal subtype with poor outcome

Prostate cancer is a biologically heterogeneous disease with variable molecular alterations underlying cancer initiation and progression. Despite recent advances in understanding prostate cancer heterogeneity, better methods for classification of prostate cancer are still needed to improve prognostic accuracy and therapeutic outcomes. In this study, we computationally assembled a large virtual cohort (n = 1,321) of human prostate cancer transcriptome profiles from 38 distinct cohorts and, using pathway activation signatures of known relevance to prostate cancer, developed a novel classification system consisting of three distinct subtypes (named PCS1-3). We validated this subtyping scheme in 10 independent patient cohorts and 19 laboratory models of prostate cancer, including cell lines and genetically engineered mouse models. Analysis of subtype-specific gene expression patterns in independent datasets derived from luminal and basal cell models provides evidence that PCS1 and PCS2 tumors reflect luminal subtypes, while PCS3 represents a basal subtype. We show that PCS1 tumors progress more rapidly to metastatic disease in comparison with PCS2 or PCS3, including PSC1 tumors of low Gleason grade. To apply this finding clinically, we developed a 37-gene panel that accurately assigns individual tumors to one of the three PCS subtypes. This panel was also applied to circulating tumor cells (CTC) and provided evidence that PCS1 CTCs may reflect enzalutamide resistance. In summary, PCS subtyping may improve accuracy in predicting the likelihood of clinical progression and permit treatment stratification at early and late disease stages. Cancer Res; 76(17); 4948-58. ©2016 AACR.

Cyclin D1 Loss Distinguishes Prostatic Small Cell Carcinoma from Most Prostatic Adenocarcinomas

Purpose: Small-cell neuroendocrine differentiation in prostatic carcinoma is an increasingly common resistance mechanism to potent androgen deprivation therapy (ADT), but can be difficult to identify morphologically. We investigated whether cyclin D1 and p16 expression can inform on Rb functional status and distinguish small-cell carcinoma from adenocarcinoma. Experimental Design: We used gene expression data and immunohistochemistry to examine cyclin D1 and p16 levels in patient-derived xenografts (PDX), and prostatic small-cell carcinoma and adenocarcinoma specimens. Results: Using PDX, we show proof-of-concept that a high ratio of p16 to cyclin D1 gene expression reflects underlying Rb functional loss and distinguishes morphologically identified small-cell carcinoma from prostatic adenocarcinoma in patient specimens (n = 13 and 9, respectively). At the protein level, cyclin D1, but not p16, was useful to distinguish small-cell carcinoma from adenocarcinoma. Overall, 88% (36/41) of small-cell carcinomas showed cyclin D1 loss by immunostaining compared with 2% (2/94) of Gleason score 7–10 primary adenocarcinomas at radical prostatectomy, 9% (4/44) of Gleason score 9–10 primary adenocarcinomas at needle biopsy, and 7% (8/115) of individual metastases from 39 patients at autopsy. Though rare adenocarcinomas showed cyclin D1 loss, many of these were associated with clinical features of small-cell carcinoma, and in a cohort of men treated with adjuvant ADT who developed metastasis, lower cyclin D1 gene expression was associated with more rapid onset of metastasis and death. Conclusions: Cyclin D1 loss identifies prostate tumors with small-cell differentiation and may identify a small subset of adenocarcinomas with poor prognosis. Clin Cancer Res; 21(24); 5619–29. ©2015 AACR.

Influence of Prior Knowledge in Constraint-Based Learning of Gene Regulatory Networks

Constraint-based structure learning algorithms generally perform well on sparse graphs. Although sparsity is not uncommon, there are some domains where the underlying graph can have some dense regions; one of these domains is gene regulatory networks, which is the main motivation to undertake the study described in this paper. We propose a new constraint-based algorithm that can both increase the quality of output and decrease the computational requirements for learning the structure of gene regulatory networks. The algorithm is based on and extends the PC algorithm. Two different types of information are derived from the prior knowledge; one is the probability of existence of edges, and the other is the nodes that seem to be dependent on a large number of nodes compared to other nodes in the graph. Also a new method based on Gene Ontology for gene regulatory network validation is proposed. We demonstrate the applicability and effectiveness of the proposed algorithms on both synthetic and real data sets.