Academic Positions

  • present 2018

    Associate Professor (tenured)

    Dept. of Biostatistics, The University of Kansas Medical Center, KUMC, Kansas City

    Member: The University of Kansas Cancer Center.

  • 2018 2014

    Post Doctoral Fellow Researcher

    Dept. of Biostatistics, MD Anderson Cancer Center, University of Texas (Houston)

    Supervisor: Prof. Ziding Feng.

Education & Training

  • Military Service 2014

    Hellenic Army

    Artillery

  • Ph.D. 2013

    Ph.D. in Statistics

    Title: "Statistical methods for the evaluation of diagnostic biomarkers in the presence of censoring".

    Supervisor: Dr. John V. Tsimikas.

    University of the Aegean, Dept. of Statistics and Actuarial-Financial Mathematics, Division of Statistics and Data Analysis.

  • M.Sc.2006

    Master in Statistics and Data Analysis

    Thesis title: "Properties of the ROC surfaces".

    Supervisor: Dr. Christos T. Nakas

    University of the Aegean, Dept. of Statistics and Actuarial-Financial Mathematics, Division of Statistics and Data Analysis.

  • Ptychion (4 year degree)2001-2005

    Statistics and Insurance Science

    University of Piraeus

My network

Software/code for implementation of published methods

  • Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.

    Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.

    Lorem ipsum dolor sit amet, consectetur adipisicing elit, sed do eiusmod tempor incididunt ut labore et dolore magna aliqua. Ut enim ad minim veniam, quis nostrud exercitation ullamco laboris nisi ut aliquip ex ea commodo consequat. Duis aute irure dolor in reprehenderit in voluptate velit esse cillum dolore eu fugiat nulla pariatur. Excepteur sint occaecat cupidatat non proident, sunt in culpa qui officia deserunt mollit anim id est laborum.

  • Statistical Inference for Box-Cox based Receiver Operating Characteristic Curves.

    It applies all approaches of the corresponding paper authored by Leonidas E. Bantis, Benjamin Brewer, Christos T. Nakas, and Benjamin Reiser

    We provide a tutorial for inferences related to Box-Cox-based ROC curves and provide a new R package named 'rocbc' that can be found on CRAN. Any use of this software implies reference to the corresponding paper. Here we provide some more detailed simulation results and the COVID19 data used in our application section. Click here to download.

  • Cutoff estimation and construction of their confidence intervals for continuous biomarkers under ternary umbrella and tree stochastic ordering settings (Codes provided by Benjamin Brewer)

    It applies all approaches of the corresponding paper authored by Benjamin Brewer and Leonidas E. Bantis. (All codes are provided by Benjamin Brewer)

    We provide new methods related to cutoff estimation for ternary tree and umbrella stochastic ordering settings. Any use of this software implies reference to the corresponding paper. Click here to download.

  • Statistical Inference for the Utility of a Biomarker based on the Partial Volume under the ROC Surface (Codes provided by Kate Young (LAST UPDATE: April 17 2023))

    It applies all approaches of the corresponding paper authored by Kate Young and Leonidas E. Bantis. (All codes are provided by Kate Young)

    We provide new methods related to pVUS estimation and inference (provides hypothesis testing and confidence intervals) in 3-class settings. Any use of this software implies reference to the corresponding paper. Click here to download.

  • Construction of joint spaces for the optimal true class fraction triplet in the ROC space using alternative biomarker cutoffs

    It applies all approaches of the corresponding paper authored by Peng Shi and Leonidas E. Bantis

    We provide new methods related to inferences around the optimal true class fraction triplet. Any use of this software implies reference to the corresponding paper. Click here to download.

  • Optimal Cutoffs and Inferences Around the True and False Classification Rates in Diagnostic Trilemmas with Applications to Pancreatic Cancer

    It applies all approaches of the corresponding paper authored by Leonidas E. Bantis and John V. Tsimikas

    We provide new methods related to inferences around the true and false class rates in trichotomous settings. Any use of this software implies reference to the corresponding paper. Click here to download.

  • Estimation and Construction of Confidence Intervals for the Cutoff-Points of Continuous Biomarkers Under the Euclidean Distance in 3-class Settings (Codes provided by Brian R. Mosier (LAST UPDATE: JANUARY 29 2022))

    It applies all approaches of the corresponding paper authored by Brian R. Mosier and Leonidas E. Bantis.

    We provide new methods related to cutoff estimation for continuous biomrkarkers in 3-class settings that are alternative to the Youden-based ones. Any use of this software implies reference to the corresponding paper. Click here to download.

  • The Length of the ROC Curve and the Two Cutoff Youden Index within a Robust Framework for Discovery, Evaluation, and Cutoff Estimation in Biomarker Studies Involving Improper ROC Curves

    It applies all approaches of the corresponding paper authored by Leonidas E. Bantis, John V. Tsimikas, Gregory Chambers, Michela Capello, Samir Hanash and Ziding Feng.

    We provide new methods related to improper ROCs. Any use of this software implies reference to the corresponding paper. Click here to download.

  • Statistical Inference for the Difference of two Youden Indices Obtained from Correlated Biomarkers

    It applies all approaches of the corresponding paper authored by Leonidas E. Bantis, Christos T. Nakas, and Benjamin Reiser.

    We provide new methods for estimating a survival function. Any use of this software implies reference to the corresponding paper. Click here to download.

  • Survival Estimation using the HCNS approach

    It applies all approaches of the corresponding paper authored by Leonidas E. Bantis, John V. Tsimikas, and Stelios Georgiou.

    We provide new methods for estimating a survival function. Any use of this software implies reference to the corresponding paper. Click here to download.

  • Estimation of Smooth ROC curves for Biomarkers With Limits of Detection

    It applies all approaches of the paper with the same title by Leonidas E. Bantis, Qingxiang Yan, John V. Tsimikas, and Ziding Feng.

    We provide new methods for estimating an ROC curve that corresponds to a biomarker that is subject to a lower and/or an upper limit of detection. We explore parametric, spline based, and kernel hybrid approaches based on multiple imputation. Any use of this software implies reference to the aforementioned paper. Click here to download.

  • Construction of confidence intervals for the maximum of the Youden index and its corresponding optimal cutoff point

    It applies all approaches of the paper with the same title by Leonidas E. Bantis, Christos T. Nakas and Benjamin Reiser.

    We provide new methods for constructing confidence intervals for both the Youden index and its corresponding cutoff point. We explore approaches based on the delta approximation under the normality assumption, as well as power transformations to normality and nonparametric kernel and spline based approaches. Any use of this software implies reference to the aforementioned paper. Click here to download.

  • Comparison of two correlated ROC surfaces for given pairs of true classification rates

    It applies all approaches of the paper with the same title by Leonidas E. Bantis, and Ziding Feng.

    Any use of this software implies reference to the corresponding paper. Click here to download.

  • image

    Restricted cubic spline

    Fits the so called restricted cubic spline via least squares and obtains 95% bootstrap based CIs.

    Fits the so called restricted cubic spline via least squares (see Harrell (2001)). The obtained spline is linear beyond the first and the last knot. The truncated power basis representation is used. That is, the fitted spline is of the form: f(x)=b0+b1*x+b2*(x-t1)^3*(x>t1)+b3*(x-t2)^3*(x>t2)+... where t1 t2,... are the desired knots. 95% confidence intervals are provided based on the bootstrap procedure. For more information see also: Frank E Harrell Jr, Regression Modelling Strategies (With application to linear models, logistic regression and survival analysis), 2001, Springer Series in Statistics, pages 20-21. Click here for download and description.
  • image

    Parametric ROC curve

    Constructs the parametric ROC curve based on parametric choices provided by the user.

    Constructs the parametric ROC curve based on parametric choices provided by the user. Estimation is done via maximum likelihood. The area under the curve (AUC) is also computed. If requested, a partial area under the curve can also be obtained. The empirical (non-parametric) ROC is also provided. Bootstrap based 95% intervals can be obtained for the corresponding AUC, the partial AUC, as well as the empirical based AUC. 95% pointwise confidence intervals can also be obtained for the parametric ROC curve itself. An informative plot depending on the choices of the user is provided automatically. The following parametric models are supported: Normal, Gamma, Lognormal and Weibull with all their possible pairwise combinations.Click here for download and description.
  • Kaplan Meier for right and/or left and/or interval censored data

    Provides the Product Limit Estimator (Kaplan Meier) for left/right/interval censored data through R.

    This works for both 32bit and 64bit machines. Please try to also run the demo included in the MATLAB R-link package, and make sure that everything works before trying this code. After that, you must download and install the packages "survival" and "matlab" in R. The analysis provided by this code is in fact the output as provided by the corresponding survival package (survfit) provided by prof. Terry Therneau. Click here for download and description.

  • Fit distributions to censored data.

    Fits a distribution to the data x based on maximum likelihood. The data can be left and/or right and/or interval censored.

    The following distributions are supported:1. Normal (normfitc) 2. Log-Normal (lognfitc) 3. Logistic (logistfitc) 4. Log-logistic (loglogistfitc) 5. Extreme value (evfitc) 6. Weibull (wblfitc) 7. Exponential (expfitc) 8. Gamma (gamfitc) 9. Rayleigh (raylfitc). Click here for download and description.

  • Accelerated Failure Time Models

    Fits the so called AFT models in the presence of right and/or left censoring..

    The “aft” function fits models of the form: Y=log(T)=g0+g1*Z1+g2*Z2+...+sigma*epsilon where usually T is a time to event variable and g0, g1, ... and sigma are to be estimated. Since T is a time to event variable censoring might be involved. The “aft” function deals with possibly right and/or left censored data. With "sigma" we denote the scale parameter, and the regression coefficients are denoted by vector g=[g0 g1 g2...]. The covariates are denoted with Z1, Z2, ... The distribution of "epsilon" defines the distribution of T. The user can specify this distribution using one of the following available options: Exponential, Weibull, Log-normal, Log-logistic, Generalized Gamma. The “aft” routine is supposed to be a MATLAB alternative to proc lifereg of SAS, or survreg of R. However the “aft” has less options. Click here for download and description.

Filter by type:

Sort by year:

Sequential Validation of Blood-Based Protein Biomarker Candidates for Early-Stage Pancreatic Cancer

Michela Capello, Leonidas E. Bantis, Ghislaine Scelo, Yang Zhao, Peng Li, Dilsher S. Dhillon, Nikul J. Patel, Deepali L. Kundnani, Hong Wang, James L. Abbruzzese, Anirban Maitra, Margaret A. Tempero, Randall Brand, Lenka Brennan, Ziding Feng, Ayumu Taguchi, Vladimir Janout, Matthew A. Firpo, Sean J. Mulvihill, Matthew H. Katz, Samir M. Hanash
Journal Paper Journal of the National Caner Institute. In press.

Abstract

Background: CA19-9, which is currently in clinical use as a pancreatic ductal adenocarcinoma (PDAC) biomarker, has limited performance in detecting early-stage disease. We and others have identified protein biomarker candidates that have the potential to complement CA19-9. We have carried out sequential validations starting with 17 protein biomarker candidates to determine which markers and marker combination would improve detection of early-stage disease compared with CA19-9 alone. Methods: Candidate biomarkers were subjected to enzyme-linked immunosorbent assay based sequential validation using independent multiple sample cohorts consisting of PDAC cases (n = 187), benign pancreatic disease (n = 93), and healthy controls (n = 169). A biomarker panel for early-stage PDAC was developed based on a logistic regression model. All statistical tests for the results presented below were one-sided. Results: Six out of the 17 biomarker candidates and CA19-9 were validated in a sample set consisting of 75 PDAC patients, 27 healthy subjects, and 19 chronic pancreatitis patients. A second independent set of 73 early-stage PDAC patients, 60 healthy subjects, and 74 benign pancreatic disease patients (combined validation set) yielded a model that consisted of TIMP1, LRG1, and CA19-9. Additional blinded testing of the model was done using an independent set of plasma samples from 39 resectable PDAC patients and 82 matched healthy subjects (test set). The model yielded areas under the curve (AUCs) of 0.949 (95% confidence interval [CI] = 0.917 to 0.981) and 0.887 (95% CI = 0.817 to 0.957) with sensitivities of 0.849 and 0.667 at 95% specificity in discriminating early-stage PDAC vs healthy subjects in the combined validation and test sets, respectively. The performance of the biomarker panel was statistically significantly improved compared with CA19-9 alone (P smaller than .001, combined validation set; P smaller than .008, test set). Conclusion: The addition of TIMP1 and LRG1 immunoassays to CA19-9 statistically significantly improves the detection of early-stage PDAC.

Quantitative imaging to evaluate malignant potential of IPMNs

Alexander N Hanania, Leonidas E. Bantis, Ziding Feng, Huamin Wang, Eric P Tamm, Matthew H Katz, Anirban Maitra, Eugene J Koay
Journal Paper Oncotarget (2016).

Abstract

Objective: To investigate using quantitative imaging to assess the malignant potential of intraductal papillary mucinous neoplasms (IPMNs) in the pancreas. Background: Pancreatic cysts are identified in over 2% of the population and a subset of these, including intraductal papillary mucinous neoplasms (IPMNs), represent pre-malignant lesions. Unfortunately, clinicians cannot accurately predict which of these lesions are likely to progress to pancreatic ductal adenocarcinoma (PDAC). Methods: We investigated 360 imaging features within the domains of intensity, texture and shape using pancreatic protocol CT images in 53 patients diagnosed with IPMN (34 “high-grade” [HG] and 19 “low-grade” [LG]) who subsequently underwent surgical resection. We evaluated the performance of these features as well as the Fukuoka criteria for pancreatic cyst resection. Results: In our cohort, the Fukuoka criteria had a false positive rate of 36%. We identified 14 imaging biomarkers within Gray-Level Co-Occurrence Matrix (GLCM) that predicted histopathological grade within cyst contours. The most predictive marker differentiated LG and HG lesions with an area under the curve (AUC) of .82 at a sensitivity of 85% and specificity of 68%. Using a cross-validated design, the best logistic regression yielded an AUC of 0.96 (σ = .05) at a sensitivity of 97% and specificity of 88%. Based on the principal component analysis, HG IPMNs demonstrated a pattern of separation from LG IPMNs. Conclusions: HG IPMNs appear to have distinct imaging properties. Further validation of these findings may address a major clinical need in this population by identifying those most likely to benefit from surgical resection

Comparison of two correlated ROC curves at a given specificity or sensitivity level

Leonidas E. Bantis, and Ziding Feng
Journal Paper Statistics in Medicine. Volume 35, Issue 24, 2016, Pages 4352–4367.

Abstract

The receiver operating characteristic (ROC) curve is the most popular statistical tool for evaluating the discriminatory capability of a given continuous biomarker. The need to compare two correlated ROC curves arises when individuals are measured with two biomarkers, which induces paired and thus correlated measurements. Many researchers have focused on comparing two correlated ROC curves in terms of the area under the curve (AUC), which summarizes the overall performance of the marker. However, particular values of specificity may be of interest. We focus on comparing two correlated ROC curves at a given specificity level. We propose parametric approaches, transformations to normality, and nonparametric kernel-based approaches. Our methods can be straightforwardly extended for inference in terms of the inverse ROC(t). This is of particular interest for comparing the accuracy of two correlated biomarkers at a given sensitivity level. Extensions also involve inference for the AUC and accommodating covariates. We evaluate the robustness of our techniques through simulations, compare them with other known approaches, and present a real-data application involving prostate cancer screening.

Construction of joint confidence regions for the optimal true class fractions of Receiver Operating Characteristic (ROC) surfaces and manifolds.

Leonidas E. Bantis, Christos T. Nakas and Benjamin Reiser.
Journal Paper Statistical Methods in Medical Research, Volume 0, Issue 0, pages 1–14, 2015.

Abstract

The three-class approach is used for progressive disorders when clinicians and researchers want to diagnose or classify subjects as members of one of three ordered categories based on a continuous diagnostic marker. The decision thresholds or optimal cut-off points required for this classification are often chosen to maximize the generalized Youden index (Nakas et al., Stat Med 2013; 32: 995–1003). The effectiveness of these chosen cut-off points can be evaluated by estimating their corresponding true class fractions and their associated confidence regions. Recently, in the two-class case, parametric and non-parametric methods were investigated for the construction of confidence regions for the pair of the Youden-index-based optimal sensitivity and specificity fractions that can take into account the correlation introduced between sensitivity and specificity when the optimal cut-off point is estimated from the data (Bantis et al., Biomet 2014; 70: 212–223). A parametric approach based on the Box–Cox transformation to normality often works well while for markers having more complex distributions a non-parametric procedure using logspline density estimation can be used instead. The true class fractions that correspond to the optimal cut-off points estimated by the generalized Youden index are correlated similarly to the two-class case. In this article, we generalize these methods to the three- and to the general k-class case which involves the classification of subjects into three or more ordered categories, where ROC surface or ROC manifold methodology, respectively, is typically employed for the evaluation of the discriminatory capacity of a diagnostic marker. We obtain three- and multi-dimensional joint confidence regions for the optimal true class fractions. We illustrate this with an application to the Trail Making Test Part A that has been used to characterize cognitive impairment in patients with Parkinson’s disease.

Construction of Confidence Regions in the ROC space After the Estimation of the Optimal Youden Index-Based Cut-off Point.

Leonidas E. Bantis, Christos T. Nakas and Benjamin Reiser.
Journal Paper Biometrics, Volume 70, Issue 1, pages 212–223, March 2014.

Abstract

After establishing the utility of a continuous diagnostic marker investigators will typically address the question of determining a cut-off point which will be used for diagnostic purposes in clinical decision making. The most commonly used optimality criterion for cut-off point selection in the context of ROC curve analysis is the maximum of the Youden index. The pair of sensitivity and specificity proportions that correspond to the Youden index-based cut-off point characterize the performance of the diagnostic marker. Confidence intervals for sensitivity and specificity are routinely estimated based on the assumption that sensitivity and specificity are independent binomial proportions as they arise from the independent populations of diseased and healthy subjects, respectively. The Youden index-based cut-off point is estimated from the data and as such the resulting sensitivity and specificity proportions are in fact correlated. This correlation needs to be taken into account in order to calculate confidence intervals that result in the anticipated coverage. In this article we study parametric and non-parametric approaches for the construction of confidence intervals for the pair of sensitivity and specificity proportions that correspond to the Youden index-based optimal cut-off point. These approaches result in the anticipated coverage under different scenarios for the distributions of the healthy and diseased subjects. We find that a parametric approach based on a Box–Cox transformation to normality often works well. For biomarkers following more complex distributions a non-parametric procedure using logspline density estimation can be used.

Comparison of two correlated ROC curves at a given sensitivity or specificity level

Leonidas Bantis, and Ziding Feng
Poster presentation ENAR-IBS 2016, Austin Texas.

Abstract

N/A

Comparison of two correlated ROC curves at a given false positive rate

Leonidas Bantis, and Ziding Feng
Oral presentation International SOciety for Clinical Biostatistics, Utrecht, 2015.

Abstract

N/A

On the accuracy of a binary time dependent biomarker

Leonidas Bantis, John V. Tsimikas, Stelios D georgiou
Conference Papers 21st pan-hellenic conference in Statistics, 2008, Samos island, Greece.

Abstract

N/A

Smooth ROC Curves and Surfaces for Markers Subject to a Limit of Detection Using Monotone Natural Cubic Splines

Leonidas E. Bantis, John V. Tsimikas, Stelios D. Georgiou
Journal Paper Biometrical Journal. Volume 55, Issue 5, pages 719–740, September 2013

Abstract

The use of ROC curves in evaluating a continuous or ordinal biomarker for the discrimination of two populations is commonplace. However, in many settings, marker measurements above or below a certain value cannot be obtained. In this paper, we study the construction of a smooth ROC curve (or surface in the case of three populations) when there is a lower or upper limit of detection. We propose the use of spline models that incorporate monotonicity constraints for the cumulative hazard function of the marker distribution. The proposed technique is computationally stable and simulation results showed a satisfactory performance. Other observed covariates can be also accommodated by this spline-based approach.

Generalized Linear Models with a Censored Covariate

Leonidas Bantis, John V. Tsimikas, Stelios D georgiou
Conference Papers 23rd pan-hellenic conference in Statistics, 2010, Veria, Greece.

Abstract

N/A

Survival Estimation with Monotone Natural Cubic Splines

Leonidas Bantis, John V. Tsimikas, Stelios D georgiou
Conference Papers 24th pan-hellenic conference in Statistics, 2011, Patra, Greece.

Abstract

N/A

Inference in GLMs with a censored covariate

Leonidas Bantis, John V. Tsimikas, Stelios D georgiou
Conference Papers 6th EMR-International Biometric Society conference, 2011, Crete, Greece.

Abstract

N/A

A MATLAB Package for Survival Estimation Using COnstrined Natural Cubic Splines

Leonidas Bantis, John V. Tsimikas, Stelios D georgiou
Conference Papers 25th pan-hellenic conference in Statistics, 2012, Volos.

Abstract

N/A

Smooth ROC curves and Surfaces for Biomarkers with a Limit of Detection.

Leonidas Bantis, John V. Tsimikas, Stelios D georgiou
Conference Papers 25th pan-hellenic conference in Statistics, 2012, Volos.

Abstract

N/A

Generalized Linear Models With a Censored Covariate for Longitudinal Data.

Leonidas Bantis, John V. Tsimikas
Conference Papers 26th pan-hellenic conference in Statistics, 2013, Pireaus.

Abstract

N/A

Construction of confidence regions in the ROC space after the estimation of the optimal Youden index-based cut-off point.

Leonidas Bantis, Christos T. Nakas, Benjamin Reiser.
Conference Papers 6th EMR-International Biometric Society conference, 2013, Tel Aviv, Israel.

Abstract

N/A

Construction of joint confidence regions for the optimal true class fractions of ROC surfaces and manifolds.

Leonidas Bantis, Christos T. Nakas, Benjamin Reiser.
Conference Papers 33rd annual conference of the International Society for Clinical Biostatistics 2014, Florence, Italy.

Abstract

N/A

Inference in Generalized Linear Regression Models with a Censored Covariate

John V. Tsimikas, Leonidas E. Bantis, Stelios D. Georgiou
Journal Paper Computational Statistics and Data Analysis. Volume 56, Issue 6, June 2012, Pages 1854–1868.

Abstract

The problem of estimating the parameters in a generalized linear model when a covariate is subject to censoring is studied. A new method based on an estimating function approach is proposed. The method does not assume a parametric form for the distribution of the response given the regressors and is computationally simple. In the linear regression case, the proposed approach implies the use of mean imputation of the censored regressor. The use of flexible parametric models for the distribution of the covariate is employed. When survival time is considered as the covariate subject to censoring, the use of the generalized gamma distribution is explored, since it is considered as a platform distribution covering a wide variety of hazard rate shapes. The method can be further robustified by considering models of nonparametric nature typically used in survival analysis such as the logspline for the censored covariate. For models involving additional, fully observed, covariates the use of a generalized gamma accelerated failure time regression model is explored. In this setting, no parametric family assumption for the extra covariates is needed. The proposed approach is broader than likelihood based multiple imputation techniques. Moreover, even in cases with a known parametric form for the response distribution, the method can be considered a feasible alternative to likelihood based estimation. Simulation studies are conducted for continuous, binary and count data to evaluate the performance of the proposed method and to compare the estimates to standard ones. An application using a well known data set of a randomized placebo controlled trial of the drug D-penicillamine (DPCA) for the treatment of primary biliary cirrhosis (PBC) conducted at the Mayo Clinic is presented. Possible extensions of the method regarding the robustness as well as the type of censoring are also discussed.

Survival estimation through the cumulative hazard function with constrained natural cubic splines

Leonidas E. Bantis, John V. Tsimikas, Stelios D. Georgiou
Journal Paper Lifetime Data Analysis. Volume 18, Issue 3, July 2012, Pages 364–396.

Abstract

In this paper we explore the estimation of survival probabilities via a smoothed version of the survival function, in the presence of censoring. We investigate the fit of a natural cubic spline on the cumulative hazard function under appropriate constraints. Under the proposed technique the problem reduces to a restricted least squares one, leading to convex optimization. The approach taken in this paper is evaluated and compared via simulations to other known methods such as the Kaplan Meier and the logspline estimator. Our approach is easily extended to address estimation of survival probabilities in the presence of covariates when the proportional hazards model assumption holds. In this case the method is compared to a restricted cubic spline approach that involves maximum likelihood. The proposed approach can be also adjusted to accommodate left censoring.

Rapid effects of humidity acclimation on stress resistance in Drosophila melanogaster

Dau Dayal Aggarwal, Poonam Rangab, Bhawna Kalrac, Ravi Parkashb, Eugenia Rashkovetskya, Leonidas E. Bantis
Journal Paper Comparative Biochemistry and Physiology Part A: Molecular & Integrative Physiology. Volume 166, Issue 1, September 2013, Pages 81–90.

Abstract

We tested the hypothesis whether developmental acclimation at ecologically relevant humidity regimes (40% and 75% RH) affects desiccation resistance of pre-adults (3rd instar larvae) and adults of Drosophila melanogaster Meigen (Diptera: Drosophilidae). Additionally, we untangled whether drought (40% RH) acclimation affects cold-tolerance in the adults of D. melanogaster. We observed that low humidity (40% RH) acclimated individuals survived significantly longer (1.6-fold) under lethal levels of desiccation stress (0–5% RH) than their counter-replicates acclimated at 75% RH. In contrast to a faster duration of development of 1st and 2nd instar larvae, 3rd instar larvae showed a delayed development at 40% RH as compared to their counterparts grown at 75% RH. Rearing to low humidity conferred an increase in bulk water, hemolymph content and dehydration tolerance, consistent with increase in desiccation resistance for replicates grown at 40% as compared to their counterparts at 75% RH. Further, we found a trade-off between the levels of carbohydrates and body lipid reserves at 40% and 75% RH. Higher levels of carbohydrates sustained longer survival under desiccation stress for individuals developed at 40% RH than their congeners at 75% RH. However, the rate of carbohydrate utilization did not differ between the individuals reared at these contrasting humidity regimes. Interestingly, our results of accelerated failure time (AFT) models showed substantial decreased death rates at a series of low temperatures (0, − 2, or − 4 °C) for replicates acclimated at 40% RH as compared to their counter-parts at 75% RH. Therefore, our findings indicate that development to low humidity conditions constrained on multiple physiological mechanisms of water-balance, and conferred cross-tolerance towards desiccation and cold stress in D. melanogaster. Finally, we suggest that the ability of generalist Drosophila species to tolerate fluctuations in humidity might aid in their existence and abundance under expected changes in moisture level in course of global climate change.

Spline based ROC curves and surfaces for biomarkers with an upper or a lower limit of detection

Leonidas E. Bantis, John V. Tsimikas, and Stelios D. Georgiou
Invited Talk 7th International Workshop on Simulation, May 21-25 (2013), Department of Statistical Sciences, University of Bologna, Italy. (session organized by Prof. Ilia Vonta).