PubMed Health⌕ Search

Biomedical subjects

Ziding Feng

Publications and source records attributed to Ziding Feng.

At least 19 recordsLinked to original sources
PubMed2024

Value of HCC surveillance in a landscape of emerging surveillance options: Perspectives of a multi-stakeholder modified Delphi panel.

HCC surveillance is recommended by liver professional societies but lacks broad acceptance by several primary care and cancer societies due to limitations in the existing data. We convened a diverse multidisciplinary group of cancer screening experts to evaluate current and future paradigms of HCC prevention and early detection using a rigorous Delphi panel approach. The experts had high agreement on 21 statements about primary prevention, HCC surveillance benefits, HCC surveillance harms, and the evaluation of emerging surveillance modalities. The experts agreed that current data have methodologic limitations as well as unclear generalizability to Western populations. Although a randomized clinical trial of surveillance versus no surveillance is unlikely feasible, they concurred that alternative designs, such as a comparison of 2 surveillance modalities, could provide indirect evidence of surveillance efficacy. The panel acknowledged the presence of surveillance harms, but concurred the overall value of surveillance appears high, particularly given a greater emphasis on benefits over harms by both patients and clinicians. The experts underscored the importance of a framework for measuring both benefits and harms when evaluating emerging surveillance strategies. The panel acknowledged performance metrics of emerging methods may differ from other cancer screening programs given differences in populations, including higher risk of cancer development and competing risk of morality, and differences in diagnostic workflow in patients at risk of HCC. These data provide insights into the perceived value of HCC surveillance in an era of emerging blood- and imaging-based surveillance strategies.

Humans↗
PubMed2005

Inactivation of p16, RUNX3, and HPP1 occurs early in Barrett's-associated neoplastic progression and predicts progression risk.

Patients with Barrett's esophagus (BE) are at increased risk of developing esophageal adenocarcinoma (EAC). Clinical neoplastic progression risk factors, such as age and the length of the esophageal BE segment, have been identified. However, improved molecular biomarkers predicting increased progression risk are needed for improved risk assessment and stratification. Using real-time quantitative methylation-specific PCR, we screened 10 genes (HPP1, RUNX3, RIZ1, CRBP1, 3-OST-2, APC, TIMP3, p16, MGMT, p14) for promoter hypermethylation in 77 EAC, 93 BE, and 64 normal esophagus (NE) specimens. A subset of genes manifesting significant differences in methylation frequencies between BE and EAC was then analysed in 20 dysplastic specimens. All 10 genes except p14 were frequently methylated in EACs, with RUNX3, HPP1, CRBP1, RIZ1, and OST-2 representing novel methylation targets in EAC and/or BE. p16, RUNX3, and HPP1 displayed increasing methylation frequencies in BE vs EAC. Furthermore, these increases in methylation occurred early, at the interface between BE and low-grade dysplasia (LGD). To demonstrate the silencing effect of hypermethylation, we selected the EAC cells BIC1, in which the HPP1 promoter is natively methylated, and subjected them to 5-aza-2'-deoxycytidine (Aza-C) treatment. Real-time RT-PCR indicated increased HPP1 mRNA levels after 3 days of Aza-C treatment, as well as decreased levels of methylated HPP1 DNA. Hypermethylation of a subset of six genes (APC, TIMP3, CRBP1, p16, RUNX3, and HPP1) was then tested in a retrospective longitudinal study of 99 BE and nine LGD specimens obtained from 53 BE patients undergoing surveillance endoscopy. Only high-grade dysplasia (HGD) or EAC were defined as progression end points. Two patient groups were compared: eight progressors (P) and 45 nonprogressors (NP), using Cox proportional hazards models to determine the relative progression risks of age, BE segment length, and methylation events. Multivariate analyses revealed that only hypermethylation of p16 (odds ratio (OR) 1.74, 95% confidence interval (CI) 1.33-2.20), RUNX3 (OR 1.80, 95% CI 1.08-2.81), and HPP1 (OR 1.77, 95% CI 1.06-2.81) were independently associated with an increased risk of progression, whereas age, BE segment length, and hypermethylation of TIMP3, APC, or CRBP1 were not independent risk factors. In combined analyses, risk was detectable up to, but not earlier than, 2 years preceding neoplastic progression. Hypermethylation of p16, RUNX3, and HPP1 in BE or LGD may represent independent risk factors for the progression of BE to HGD or EAC. These findings have implications regarding risk stratification, early EAC detection, and the appropriate endoscopic surveillance interval for patients with BE.

Adenocarcinoma↗
PubMed2005

Optimized normalization for antibody microarrays and application to serum-protein profiling.

The measurements of coordinated patterns of protein abundance using antibody microarrays could be used to gain insight into disease biology and to probe the use of combinations of proteins for disease classification. The correct use and interpretation of antibody microarray data requires proper normalization of the data, which has not yet been systematically studied. Therefore we undertook a study to determine the optimal normalization of data from antibody microarray profiling of proteins in human serum specimens. Forty-three serum samples collected from patients with pancreatic cancer and from control subjects were probed in triplicate on microarrays containing 48 different antibodies, using a direct labeling, two-color comparative fluorescence detection format. Seven different normalization methods representing major classes of normalization for antibody microarray data were compared by their effects on reproducibility, accuracy, and trends in the data set. Normalization with ELISA-determined concentrations of IgM resulted in the most accurate, reproducible, and reliable data. The other normalization methods were deficient in at least one of the criteria. Multiparametric classification of the samples based on the combined measurement of seven of the proteins demonstrated the potential for increased classification accuracy compared with the use of individual measurements. This study establishes reliable normalization for antibody microarray data, criteria for assessing normalization performance, and the capability of antibody microarrays for serum-protein profiling and multiparametric sample classification.

Adenocarcinoma↗
PubMed2005

Clinical and biological features associated with epidermal growth factor receptor gene mutations in lung cancers.

BACKGROUND: Mutations in the tyrosine kinase (TK) domain of the epidermal growth factor receptor (EGFR) gene in lung cancers are associated with increased sensitivity of these cancers to drugs that inhibit EGFR kinase activity. However, the role of such mutations in the pathogenesis of lung cancers is unclear. METHODS: We sequenced exons 18-21 of the EGFR TK domain from genomic DNA isolated from 617 non-small-cell lung cancers (NSCLCs) and 524 normal lung tissue samples from the same patients and 36 neuroendocrine lung tumors collected from patients in Japan, Taiwan, the United States, and Australia and from 243 other epithelial cancers. Mutation status was compared with clinicopathologic features and with the presence of mutations in KRAS, a gene in the EGFR signaling pathway that is also frequently mutated in lung cancers. All statistical tests were two sided. RESULTS: We detected a total of 134 EGFR TK domain mutations in 130 (21%) of the 617 NSCLCs but not in any of the other carcinomas, nor in nonmalignant lung tissue from the same patients. In NSCLC patients, EGFR TK domain mutations were statistically significantly more frequent in never smokers than ever smokers (51% versus 10%), in adenocarcinomas versus cancer of other histologies (40% versus 3%), in patients of East Asian ethnicity versus other ethnicities (30% versus 8%), and in females versus males (42% versus 14%; all P < .001). EGFR TK domain mutation status was not associated with patient age at diagnosis, clinical stage, the presence of bronchioloalveolar histologic features, or overall survival. The EGFR TK domain mutations we detected were of three common types: in-frame deletions in exon 19, single missense mutations in exon 21, and in-frame duplications/insertions in exon 20. Rare missense mutations were also detected in exons 18, 20, and 21. KRAS gene mutations were present in 50 (8%) of the 617 NSCLCs but not in any tumors with an EGFR TK domain mutation. CONCLUSIONS: Mutations in either the EGFR TK domain or the KRAS gene can lead to lung cancer pathogenesis. EGFR TK domain mutations are the first molecular change known to occur specifically in never smokers.

Adenocarcinoma↗
PubMed2005

Reliability, effect size, and responsiveness of health status measures in the design of randomized and cluster-randomized trials.

BACKGROUND: New health status survey instruments are often described by their psychometric (measurement) properties, such as Validity, Reliability, Effect Size, and Responsiveness. For cluster-randomized trials, another important statistic is the Intraclass Correlation (ICC) for the instrument within clusters. Studies using better instruments can be performed with smaller sample sizes, but better instruments may be more expensive in terms of dollars, opportunity cost, or poorer data quality due to the response burden of longer instruments. METHODS: We defined the psychometric statistics in terms of a mathematical model, and examined the power of a two-sample test as a function of the test-retest Reliability, Effect Size, Responsiveness, and Intraclass Correlation of the instrument. We examined the "cost-effectiveness" of using a one-item versus a five-item measure of mental health status. FINDINGS: Under the standard model for measurement error, the psychometric statistics are all functions of the same error term. They are also functions of the setting in which they were estimated. In randomized trials, power is a function of Reliability and sample size, and a less reliable instrument can achieve the desired power if N is increased. In cluster-randomized trials, adequate power may be obtained by increasing the number of clusters per treatment group (and often the number of persons per cluster), as well as by choosing a more reliable instrument. The one-item measure of mental health status may be more cost-effective than the five-item measure in some situations. CONCLUSION: If the goal is to diagnose or refer individual patients, an instrument with high Validity and Reliability is needed. In settings where the sample sizes are large or can be increased easily, any valid instrument may be cost-effective. It is likely that many published values of psychometric statistics are accurate only in settings similar to that in which they were estimated.

Cost-Benefit Analysis↗
PubMed2005

Differential methylation of genes that regulate cytokine signaling in lymphoid and hematopoietic tumors.

The perturbations of the cytokine signaling pathway play an important role in lymphoid/hematopoietic tumors. Aberrant promoter methylation is the major mechanism of gene silencing in tumors. We examined 150 lymphoid/hematopoietic tumors or potential premalignant specimens, 55 control specimens and 12 EBV-transformed B lymphoblastoid cultures and 10 lymphoma/leukemia (L/L) or multiple myeloma (MM) cell lines for the methylation (and, in cell lines, of the expression status) of three genes involved in the cytokine signaling pathway. The genes were: SHP1, a protein tyrosine phosphatase; SYK, a protein kinase; and SOCS1, a suppressor of cytokine signaling. Our major findings were: (1) one or more of the three genes was frequently methylated in L/L and MM cell lines and there was good concordance (90-100%) between methylation and loss of gene expression; (2) treatment of L/L cell lines with a demethylating agent resulted in re-expression of SHP1 protein and downregulation of phosphorylated STAT3 in L/L cell lines; (3) all 55 control specimens and the lymphoblastoid cultures were negative for methylation of the three genes; (4) non-Hodgkin's lymphomas (100%), and leukemias (94%) had almost universal methylation of SHP1 and relatively less frequent (<30%) methylation of SOCS1 and SYK; (5) MM and monoclonal gammopathy of unknown significance (MGUS) had infrequent methylation of SHP1 (<20%), and occasional methylation of SOCS1 and SYK; and (6) comparable methylation frequencies for SOCS1 were observed in MM and MGUS, suggesting that SOCS1 methylation is an early event in MM pathogenesis. At least one gene was methylated in 119 of 130 (93%) of the malignant and 12 of 20 (60%) of the MGUS samples. Our findings demonstrate that the perturbations of cytokine signaling via silencing of these three genes are almost universal in lymphoid/hematopoietic tumors but the patterns of gene methylated for L/L and plasma cell dyscrasias are different.

Cell Line, Tumor↗
PubMed2005

5-year urinary and sexual outcomes after radical prostatectomy: results from the prostate cancer outcomes study.

PURPOSE: Prior studies of postoperative outcomes following radical prostatectomy have been limited by selection bias and short-term followup. In this study we assessed temporal changes in urinary and sexual function up to 5 years following radical prostatectomy in a population based cohort. MATERIALS AND METHODS: A sample of 1,288 men with localized prostate cancer who underwent radical prostatectomy and completed a baseline survey within 6 to 12 months of diagnosis were included in the analysis. Two and 5-year functional and quality of life data were collected, as was information on the use of erectile aids. Temporal functional changes and potentially confounding or modifying factors were assessed using longitudinal regression models. RESULTS: Of these men 14% reported frequent urinary leakage or no urinary control 60 months after diagnosis, which was slightly higher than the 10% reporting incontinence at 24 months (p = 0.007). At 60 months 28% of the men had erections firm enough for intercourse compared with 22% at 24 months (p = 0.003). Sildenafil was the most commonly used erectile aid (43% ever used) and 45% of users reported that it helped "somewhat" or "a lot." CONCLUSIONS: Urinary and sexual dysfunction were common 5 years following radical prostatectomy in this large, community based cohort of prostate cancer survivors. While a small minority of subjects experienced changes in urinary or sexual function between years 2 and 5 after prostatectomy, functional outcomes remained relatively stable in the majority of participants.

Adult↗
PubMed2005

Evaluation of serum protein profiling by surface-enhanced laser desorption/ionization time-of-flight mass spectrometry for the detection of prostate cancer: I. Assessment of platform reproducibility.

BACKGROUND: Protein expression profiling for differences indicative of early cancer has promise for improving diagnostics. This report describes the first stage of a National Cancer Institute/Early Detection Research Network-sponsored multiinstitutional evaluation and validation of this approach for detection of prostate cancer. METHODS: Two sequential experimental phases were conducted to establish interlaboratory calibration and standardization of the surface-enhanced laser desorption (SELDI) instrumental and assay platform output. We first established whether the output from multiple calibrated Protein Biosystem II SELDI-ionization time-of-flight mass spectrometry (TOF-MS) instruments demonstrated acceptable interlaboratory reproducibility. This was determined by measuring mass accuracy, resolution, signal-to-noise ratio, and normalized intensity of three m/z "peaks" present in a standard pooled serum sample. We next evaluated the ability of the calibrated and standardized instrumentation to accurately differentiate between selected cases of prostate cancer and control by use of an algorithm developed from data derived from a single site 2 years earlier. RESULTS: When the described standard operating procedures were established at all laboratory sites, the across-laboratory measurements revealed a CV for mass accuracy of 0.1%, signal-to-noise ratio of approximately 40%, and normalized intensity of 15-36% for the three pooled serum peaks. This was comparable to the intralaboratory measurements of the same peaks. The instrument systems were then challenged with sera from a selected group of 14 cases and 14 controls. The classification agreement between each site and the established decision algorithm were examined by use of both raw peak intensity boosting and ranked peak intensity boosting. All six sites achieved perfect blinded classification for all samples when boosted alignment of raw intensities was used. Four of six sites achieved perfect blinded classification with ranked intensities, with one site passing the criteria of 26 of 28 correct and one site failing with 19 of 28 correct. CONCLUSIONS: These results demonstrate that "between-laboratory" reproducibility of SELDI-TOF-MS serum profiling approaches that of "within-laboratory" reproducibility as determined by measuring discrete m/z peaks over time and across laboratories.

Calibration↗
PubMed2004

Vitamin D receptor gene polymorphisms and prostate cancer risk.

BACKGROUND: Vitamin D is a steroid hormone that is thought to play a role in the etiology and progression of prostate cancer. Hormone activity requires binding to the vitamin D receptor (VDR), which contains several genetic polymorphisms that have been associated with risk of prostate cancer. To further evaluate this relationship, we conducted a population-based case-control study of the VDR BsmI, FokI, and Poly-A polymorphisms, and prostate cancer. METHODS: Germline DNA samples and survey data from incident prostate cancer cases (n = 559) and controls (n = 523) of similar age (40-64 years) without a history of the disease who resided in King County, Washington were analyzed. RESULTS: The frequency of the BsmI, FokI, and Poly-A genotypes were similar in cases and controls, and no overall association between any variants and prostate cancer risk were noted. Stratification by clinical features of disease revealed that among men with localized stage disease, the BsmI bb genotype was associated with a modest increase in risk (OR, 1.49; 95% CI, 1.02-2.17; P = 0.04) compared to the BB genotype. CONCLUSIONS: These results suggest that polymorphisms in the VDR gene are not strong predictors of prostate cancer risk among Caucasian men in the U.S.

Adult↗
PubMed2004

Presence of simian virus 40 DNA sequences in human lymphoid and hematopoietic malignancies and their relationship to aberrant promoter methylation of multiple genes.

The simian polyoma virus SV40 has been detected in specific human tumors including non-Hodgkin's lymphomas, although a causative role for the virus has not been convincingly demonstrated. Aberrant methylation of CpG islands in promoter regions is a frequent method of silencing tumor suppressor genes (TSGs) in cancers and may be induced by oncogenic viruses. We investigated the relationship between the presence of SV40 or EBV DNA sequences and the methylation profiles for 10 TSGs in 90 cases of non-Hodgkin's lymphomas/leukemias and 56 control tissues. SV40 sequences were present in 33/90 (37%) non-Hodgkin's lymphomas/leukemias, and EBV was present in 11/42 (26%) of non-Hodgkin's lymphomas. We found a highly significant correlation between the presence of SV40 and methylation of seven genes (P values, 0.006 to <0.0001). In lymphomas, there was no relationship between EBV and methylation. Oncogenic viruses and methylation were rarely present in control tissues. We investigated methylation of the same 10 TSGs in peripheral blood mononuclear cells (PBMC) from a healthy volunteer infected with EBV or EBV and SV40. Promoter methylation of CDH1 and CDH13 were noted in dual SV40- and EBV-infected PBMC, and these two genes were also highly significantly correlated to the presence of SV40 sequences in tumors. SV40 infection also resulted in appearance of the lymphoma/leukemia-specific marker, methylated SHP1. Methylation was completely absent in uninfected and EBV-infected PBMC. Our results demonstrate that the presence of SV40 in hematological malignancies is associated with promoter methylation of TSGs and that in all probability, the virus plays a role in tumor pathogenesis.

DNA Methylation↗
PubMed2004

DNA methylation profiles of lymphoid and hematopoietic malignancies.

PURPOSE: Aberrant methylation of the 5' gene promoter regions is an epigenetic phenomenon that is the major mechanism for silencing of tumor suppressor genes in many cancer types. The aims of our study were (a) to compare the methylation profiles of the major forms of hematological malignancies and (b) to determine the methylation profile of monoclonal gammopathy of undetermined significance (MGUS) and compare it with that of multiple myeloma (MM). EXPERIMENTAL DESIGN: We compared the aberrant promoter methylation profile of 14 known or suspected tumor suppressor genes in leukemias (n = 48), lymphomas (n = 42), and MMs (n = 40). We also examined the methylation profile of MGUS (n = 20), a premalignant plasma cell dyscrasia. The genes studied represent five of the six "hallmarks of cancer." RESULTS: Peripheral blood lymphocytes (n = 14) from healthy volunteers were negative for methylation of all genes, and methylation percentages in 41 nonmalignant tissues (peripheral blood mononuclear cells, bone marrows, and lymph nodes) from hematological patients were low (0-9%) for all 14 genes, confirming that methylation was tumor specific. Ten of the genes were methylated at frequencies of 29-68% in one or more tumor types, and the methylation indices (an indicator of overall methylation) varied from 0.25 to 0.34. With two exceptions, the methylation patterns of leukemias and lymphomas were similar. However, the pattern of MMs varied from the other tumor types for six genes. In general, the methylation pattern of MGUS was similar to that of MM, although the methylation frequencies were lower (the methylation index of MGUS was 0.15, and that of MM was 0.3). However, the methylation frequencies of six genes were significantly higher in MGUS than in control tissues. The relatively high frequencies of methylation in MGUS are consistent with it being a premalignant condition. CONCLUSIONS: The three major forms of lymphoid/hematopoietic malignancies show overlapping but individual patterns of methylation.

Age Factors↗
PubMed2004

Syndecan-1 expression in locally invasive and metastatic prostate cancer.

OBJECTIVES: To determine the significance of syndecan-1 expression, a cell-surface heparan sulfate proteoglycan in localized and metastatic prostate cancer. METHODS: We performed a retrospective analysis of 76 men with Gleason sum 6 or 7 prostate cancer treated by radical prostatectomy and a separate cohort of 75 men with metastatic prostate cancer. Syndecan-1 immunoreactivity was measured in primary prostate specimens or in samples from metastatic sites and correlated with patient outcome. RESULTS: Syndecan-1 was expressed in normal basal and secretory epithelial cells, 26% of radical prostatectomy specimens, and 35% of metastatic disease. No association was found between syndecan-1 positivity and prostate-specific antigen recurrence in the collective cohort of Gleason sum 6 and 7 cancers. However, when stratified by Gleason sum, syndecan-1 immunoreactivity (immunoreactivity score 150 or greater) was associated with a greater recurrence rate in Gleason sum 7 cancers. Expression of syndecan-1 was significantly greater in soft tissue than in bone metastasis (P = 0.048, Fisher's exact test). CONCLUSIONS: Consistent with a possible biochemical role for syndecan-1 in prostate cancer progression and metastasis, syndecan-1 expression correlated with serologic recurrence in Gleason sum 7 prostate cancer and was highly expressed in soft-tissue metastases.

Adenocarcinoma↗
PubMed2004

Baseline data and design for a randomized intervention study of dietary change in religious organizations.

BACKGROUND: Reducing dietary fat has been identified as a potential means of preventing chronic disease. Several studies have identified methods of changing dietary fat consumption in small, intensive intervention settings. Fewer studies have examined how to improve dietary habits of individuals in the general public. METHODS: The Eating for a Healthy Life (EHL) project was a randomized trial [n = 40 religious organizations (ROs) and 2175 individuals, 1099 in intervention]. The study tested an intervention package of self-help books, motivational messages, and social interactions designed to change dietary behaviors (lowering fat, increasing fruit and vegetable consumption) among members of religious organizations aged 18 and over. The primary outcomes are fat- and fruit- or vegetable-related behaviors measured using the Fat and Fiber Behavior (FFB) Questionnaire. RESULTS: Religious organizations and their members participated fully in the initial study activities. CONCLUSIONS: This study will support a rigorous test of the intervention package.

Adolescent↗
PubMed2004

Partially supervised learning using an EM-boosting algorithm.

Training data in a supervised learning problem consist of the class label and its potential predictors for a set of observations. Constructing effective classifiers from training data is the goal of supervised learning. In biomedical sciences and other scientific applications, class labels may be subject to errors. We consider a setting where there are two classes but observations with labels corresponding to one of the classes may in fact be mislabeled. The application concerns the use of protein mass-spectrometry data to discriminate between serum samples from cancer and noncancer patients. The patients in the training set are classified on the basis of tissue biopsy. Although biopsy is 100% specific in the sense that a tissue that shows itself to have malignant cells is certainly cancer, it is less than 100% sensitive. Reference gold standards that are subject to this special type of misclassification due to imperfect diagnosis certainty arise in many fields. We consider the development of a supervised learning algorithm under these conditions and refer to it as partially supervised learning. Boosting is a supervised learning algorithm geared toward high-dimensional predictor data, such as those generated in protein mass-spectrometry. We propose a modification of the boosting algorithm for partially supervised learning. The proposal is to view the true class membership of the samples that are labeled with the error-prone class label as missing data, and apply an algorithm related to the EM algorithm for minimization of a loss function. To assess the usefulness of the proposed method, we artificially mislabeled a subset of samples and applied the original and EM-modified boosting (EM-Boost) algorithms for comparison. Notable improvements in misclassification rates are observed with EM-Boost.

Algorithms↗
PubMed2004

Evaluation of community-intervention trials via generalized linear mixed models.

In community-intervention trials, communities, rather than individuals, are randomized to experimental arms. Generalized linear mixed models offer a flexible parametric framework for the evaluation of community-intervention trials, incorporating both systematic and random variations at the community and individual levels. We propose here a simple two-stage inference method for generalized linear mixed models, specifically tailored to the analysis of community-intervention trials. In the first stage, community-specific random effects are estimated from individual-level data, adjusting for the effects of individual-level covariates. This reduces the model approximately to a linear mixed model with the unit of analysis being community. Because the number of communities is typically small in community-intervention studies, we apply the small-sample inference method of Kenward and Roger (1997, Biometrics53, 983-997) to the linear mixed model of second stage. We show by simulation that, under typical settings of community-intervention studies, the proposed approach improves the inference on the intervention-effect parameter uniformly over both the linearized mixed-effect approach and the adaptive Gaussian quadrature approach for generalized linear mixed models. This work is motivated by a series of large randomized trials that test community interventions for promoting cancer preventive lifestyles and behaviors.

Biometry↗
PubMed2004

Research issues and strategies for genomic and proteomic biomarker discovery and validation: a statistical perspective.

The development and validation of clinically useful biomarkers from high-dimensional genomic and proteomic information pose great research challenges. Present bottlenecks include: that few of the biomarkers showing promise in initial discovery were found to warrant subsequent validation; and biomarker validation is expensive and time consuming. Biomarker evaluation should proceed in an orderly fashion to enhance rigor and efficiency. A molecular profiling approach, although promising, has a high chance of yielding biased results and overfitted models. Specimens from cohorts or intervention trials are essential to eliminate biases. The high cost for biomarker validation motivates some novel study design features, including sequential filtering and DNA pooling. For data analysis, logistic regression (in particular, boosting logistic regression) has features of robustness against model misspecification, and has resistance to model overfitting. Model assessment and cross-validation are critical components of data analysis. Having an independent test set is a vital feature of study design.

Biomarkers, Tumor↗