1 May 2018 in EBioMedicine
Hypoxia is associated with a poor prognosis in prostate cancer. This work aimed to derive and validate a hypoxia-related mRNA signature for localized prostate cancer. Hypoxia genes were identified in vitro via RNA-sequencing and combined with in vivo gene co-expression analysis to generate a signature. The signature was independently validated in eleven prostate cancer cohorts and a bladder cancer phase III randomized trial of radiotherapy alone or with carbogen and nicotinamide (CON). A 28-gene signature was derived. Patients with high signature scores had poorer biochemical recurrence free survivals in six of eight independent cohorts of prostatectomy-treated patients (Log rank test P < .05), with borderline significances achieved in the other two (P < .1). The signature also predicted biochemical recurrence in patients receiving post-prostatectomy radiotherapy (n = 130, P = .007) or definitive radiotherapy alone (n = 248, P = .035). Lastly, the signature predicted metastasis events in a pooled cohort (n = 631, P = .002). Prognostic significance remained after adjusting for clinic-pathological factors and commercially available prognostic signatures. The signature predicted benefit from hypoxia-modifying therapy in bladder cancer patients (intervention-by-signature interaction test P = .0026), where carbogen and nicotinamide was associated with improved survival only in hypoxic tumours. A 28-gene hypoxia signature has strong and independent prognostic value for prostate cancer patients.
1 February 2016 in Expert Systems with Applications
LPLazaros G. Papageorgiou
A novel piece-wise linear regression method has been proposed in this work.The method partitions samples into multiple regions from a single attribute.Each region is fitted with a linear regression function.An optimisation model is proposed to decide break-points and regression functions.Benchmark examples have been used to demonstrate its efficiency. In data mining, regression analysis is a computational tool that predicts continuous output variables from a number of independent input variables, by approximating their complex inner relationship. A large number of methods have been successfully proposed, based on various methodologies, including linear regression, support vector regression, neural network, piece-wise regression, etc. In terms of piece-wise regression, the existing methods in literature are usually restricted to problems of very small scale, due to their inherent non-linear nature. In this work, a more efficient piece-wise linear regression method is introduced based on a novel integer linear programming formulation. The proposed method partitions one input variable into multiple mutually exclusive segments, and fits one multivariate linear regression function per segment to minimise the total absolute error. Assuming both the single partition feature and the number of regions are known, the mixed integer linear model is proposed to simultaneously determine the locations of multiple break-points and regression coefficients for each segment. Furthermore, an efficient heuristic procedure is presented to identify the key partition feature and final number of break-points. 7 real world problems covering several application domains have been used to demonstrate the efficiency of our proposed method. It is shown that our proposed piece-wise regression method can be solved to global optimality for datasets of thousands samples, which also consistently achieves higher prediction accuracy than a number of state-of-the-art regression methods. Another advantage of the proposed method is that the learned model can be conveniently expressed as a small number of if-then rules that are easily interpretable. Overall, this work proposes an efficient rule-based multivariate regression method based on piece-wise functions and achieves better prediction performance than state-of-the-arts approaches. This novel method can benefit expert systems in various applications by automatically acquiring knowledge from databases to improve the quality of knowledge base.
1 July 2017 in Expert Systems with Applications
LPLazaros G. Papageorgiou
This work proposes a novel tree model for multivariate regression analysis.Both node splitting and regression coefficients are optimised in this model.The proposed method achieves improved prediction accuracy than literature methods.The resultant trees are at least as simple as the ones from previous methods. Regression analysis is a machine learning approach that aims to accurately predict the value of continuous output variables from certain independent input variables, via automatic estimation of their latent relationship from data. Tree-based regression models are popular in literature due to their flexibility to model higher order non-linearity and great interpretability. Conventionally, regression tree models are trained in a two-stage procedure, i.e. recursive binary partitioning is employed to produce a tree structure, followed by a pruning process of removing insignificant leaves, with the possibility of assigning multivariate functions to terminal leaves to improve generalisation. This work introduces a novel methodology of node partitioning which, in a single optimisation model, simultaneously performs the two tasks of identifying the break-point of a binary split and assignment of multivariate functions to either leaf, thus leading to an efficient regression tree model. Using six real world benchmark problems, we demonstrate that the proposed method consistently outperforms a number of state-of-the-art regression tree models and methods based on other techniques, with an average improvement of 760% on the mean absolute errors (MAE) of the predictions.
14 August 2017 in Clinical Cancer Research
Purpose: Hypoxia modification improves overall survival in muscle-invasive bladder cancer patients who undergo radiotherapy. There is evidence that hypoxic tumors benefit most from hypoxia modification. The study aimed to identify or derive a hypoxia gene signature that predicts benefit from hypoxia-modifying treatment in bladder cancer. Experimental Design: Published hypoxia signatures were tested and a new one derived by analyzing bladder cancer transcriptomic data from public databases. Tumor samples were available from the BCON phase III randomized trial of radiotherapy alone or with carbogen and nicotinamide (CON). Gene expression data were generated for 151 tumors using Affymetrix Human 1.0 Exon ST arrays and used for independent validation. Results: A 24-gene signature was derived, which was prognostic in four of six independent surgical cohorts (n = 679; meta HR, 2.32; 95% CI, 1.73–3.12; P < 0.0001). The signature was also prognostic in BCON patients receiving radiotherapy alone (n = 75; HR for local relapse-free survival, 2.37; 95% CI, 1.26–4.47; P = 0.0076). The signature predicted benefit from CON (n = 76; HR, 0.47; 95% CI, 0.26–0.86; P = 0.015). Prognostic significance (P = 0.017) and predictive significance (P = 0.058) remained after adjusting for clinicopathologic variables. A test for interaction between hypoxia status and treatment arms was significant (P = 0.0094). Conclusions: A 24-gene hypoxia signature has strong and independent prognostic and predictive value for muscle-invasive bladder cancer patients. The signature can aid identification of patients likely to benefit from the addition of carbogen and nicotinamide to radiotherapy. Clin Cancer Res; 23(16); 4761–8. ©2017 AACR.
12 December 2017 in Oncotarget
Validation of a hypoxia related gene signature in multiple soft tissue sarcoma cohorts Validation of a hypoxia related gene signature in multiple soft tissue sarcoma cohorts
1 March 2012 in Applied Energy
LPLazaros G. Papageorgiou