A statistical testing procedure for validating class labels.

Motivated by an open problem of validating protein identities in label-free shotgun proteomics work-flows, we present a testing procedure to validate class (protein) labels using available measurements across N instances (peptides). More generally, we present a non-parametric solution to the problem of identifying instances that are deemed as outliers relative to the subset of instances assigned to the same class. The primary assumption is that measured distances between instances within the same class are stochastically smaller than measured distances between instances from different classes. We show that the overall type I error probability across all instances within a class can be controlled by some fixed value (say α). We also demonstrate conditions where similar results on type II error probability hold. The theoretical results are supplemented by an extensive numerical study illustrating the applicability and viability of our method. Even with up to 25% of instances initially mislabeled, our testing procedure maintains a high specificity and greatly reduces the proportion of mislabeled instances. The applicability and effectiveness of our testing procedure is further illustrated by a detailed example on a proteomics data set from children with sickle cell disease where five spike-in proteins acted as contrasting controls.

Developing Drugs for Tissue-Agnostic Indications: A Paradigm Shift in Leveraging Cancer Biology for Precision Medicine.

Targeted therapies have reshaped the landscape of the development of cancer therapeutics. Recent biomarker-driven, tissue-agnostic clinical trials represent a significant paradigm shift in precision cancer medicine. Despite their growth in preclinical and clinical studies, to date only a few biomarker-driven, tissue-agnostic indications have seen approval by the US Food and Drug Administration (FDA). These approvals include pembrolizumab in microsatellite instability-high or mismatch repair deficient solid tumors, as well as both larotrectinib and entrectinib in NTRK fusion-positive tumors. Complex cancer biology, clinical trial design, and identification of resistance mechanisms represent some of the challenges that future tissue-agnostic therapies have to overcome. In this Review, we present a brief history of the development of tissue-agnostic therapies, comparing the similarities in the approval of pembrolizumab, larotrectinib, and entrectinib for tissue-agnostic indications. We also explore the future of tissue-agnostic cancer therapeutics while identifying important challenges for the future that drugs targeting tissue-agnostic indications will face.

Evidence for BCR/ABL1-positive T-cell acute lymphoblastic leukemia arising in an early lymphoid progenitor cell.

BCR-ABL1-positive leukemias have historically been classified as either chronic myelogenous leukemia or Ph+ acute lymphoblastic leukemia. Recent analyses suggest there may be a wider range of subtypes. We report a patient with BCR-ABL1 fusion positive T-cell ALL with a previously undescribed cell distribution of the fusion gene. The examination of sorted cells by fluorescence in situ hybridization showed the BCR-ABL1 fusion in the malignant T cells and a subpopulation of the nonmalignant B cells, but not nonmalignant T cells or myeloid or CD34+ progenitor cells providing evidence that the fusion may have occurred in an early lymphoid progenitor.

The Effect of Molecular Weight on Passage of Proteins Through the Blood-Aqueous Barrier.

Purpose: To determine the effect of molecular weight (MW) on the concentration of plasma-derived proteins in aqueous humor and to estimate the plasma-derived and eye-derived fractions for each protein. Methods: Aqueous humor and plasma samples were obtained during cataract surgery on an institutional review board-approved protocol. Protein concentrations were determined by ELISA and quantitative antibody microarrays. A total of 93 proteins were studied, with most proteins analyzed using 27 to 116 aqueous and 6 to 30 plasma samples. Results: Plasma proteins without evidence of intraocular expression by sequence tags were used to fit a logarithmic model relating aqueous-plasma ratio (AH:PL) to MW. The log(AH:PL) appears to be well predicted by the log(MW) (P < 0.0001), with smaller proteins such as cystatin C (13 kDa) having a higher AH:PL (1:6) than larger proteins such as albumin (66 kDa, 1:300) and complement component 5 (188 kDa, 1:2500). The logarithmic model was used to calculate the eye-derived intraocular fraction (IOF) for each protein. Based on the IOF, 66 proteins could be categorized as plasma-derived (IOF<20), whereas 10 proteins were primarily derived from eye tissue (IOF >80), and 17 proteins had contribution from both plasma and eye tissue (IOF 20-80). Conclusions: Protein concentration of plasma-derived proteins in aqueous is nonlinearly dependent on MW in favor of smaller proteins. Our study demonstrates that for proper interpretation of results, proteomic studies evaluating changes in aqueous humor protein levels should take into account the plasma and eye-derived fractions.

Analysis of serum Hsp90 as a potential biomarker of ß cell autoimmunity in type 1 diabetes.

Heat shock protein 90 (Hsp90) is a protein chaperone that is upregulated and released from pancreatic ß cells under pro-inflammatory conditions. We hypothesized that serum Hsp90 may have utility as a biomarker of type 1 diabetes risk and exhibit elevations before the onset of clinically significant hyperglycemia. To this end, total levels of the alpha cytoplasmic isoform of Hsp90 were assayed in autoantibody-positive progressors to type 1 diabetes using banked serum samples from the TrialNet Pathway to Prevention Cohort that had been collected 12 months prior to diabetes onset, with comparison to age, sex, and BMI-category matched autoantibody-positive nonprogressors and healthy controls. Hsp90 levels were higher in autoantibody-positive progressors and nonprogressors ≤ 18 years of age compared to matched healthy controls. However, Hsp90 levels were not different between progressors and nonprogressors in any age group. Hsp90 was positively correlated with age in control subjects, but this correlation was absent in autoantibody positive individuals. In aggregate these data indicate that elevated Hsp90 levels are present in youth with ß cell autoimmunity, but are not able to distinguish youth or adult type 1 diabetes progressors from nonprogressors in samples collected 12 months prior to diabetes development.

Statistical protein quantification and significance analysis in label-free LC-MS experiments with complex designs.

BACKGROUND: Liquid chromatography coupled with tandem mass spectrometry (LC-MS/MS) is widely used for quantitative proteomic investigations. The typical output of such studies is a list of identified and quantified peptides. The biological and clinical interest is, however, usually focused on quantitative conclusions at the protein level. Furthermore, many investigations ask complex biological questions by studying multiple interrelated experimental conditions. Therefore, there is a need in the field for generic statistical models to quantify protein levels even in complex study designs. RESULTS: We propose a general statistical modeling approach for protein quantification in arbitrary complex experimental designs, such as time course studies, or those involving multiple experimental factors. The approach summarizes the quantitative experimental information from all the features and all the conditions that pertain to a protein. It enables both protein significance analysis between conditions, and protein quantification in individual samples or conditions. We implement the approach in an open-source R-based software package MSstats suitable for researchers with a limited statistics and programming background. CONCLUSIONS: We demonstrate, using as examples two experimental investigations with complex designs, that a simultaneous statistical modeling of all the relevant features and conditions yields a higher sensitivity of protein significance analysis and a higher accuracy of protein quantification as compared to commonly employed alternatives. The software is available at http://www.stat.purdue.edu/~ovitek/Software.html.

Statistical design and analysis of label-free LC-MS proteomic experiments: a case study of coronary artery disease.

This chapter presents a case study, which applies statistical design and analysis to an LC-MS-based -investigation of subjects with coronary artery disease. First, we discuss the principles of statistical -experimental design, and the specification of an Analysis of Variance (ANOVA) model that describes the major sources of variation in the data. Second, we discuss procedures for detecting differentially abundant proteins, estimating protein abundance in individual samples, testing predefined groups of proteins for enrichment in differential abundance, and calculating sample size for a future experiment. The discussion is accompanied by examples of computer code implemented in the open-source statistical software R, which can be followed for an independent implementation of a similar investigation.

Identification and quantification of metabolites in (1)H NMR spectra by Bayesian model selection.

MOTIVATION: Nuclear magnetic resonance (NMR) spectroscopy is widely used for high-throughput characterization of metabolites in complex biological mixtures. However, accurate interpretation of the spectra in terms of identities and abundances of metabolites can be challenging, in particular in crowded regions with heavy peak overlap. Although a number of computational approaches for this task have recently been proposed, they are not entirely satisfactory in either accuracy or extent of automation. RESULTS: We introduce a probabilistic approach Bayesian Quantification (BQuant), for fully automated database-based identification and quantification of metabolites in local regions of (1)H NMR spectra. The approach represents the spectra as mixtures of reference profiles from a database, and infers the identities and the abundances of metabolites by Bayesian model selection. We show using a simulated dataset, a spike-in experiment and a metabolomic investigation of plasma samples that BQuant outperforms the available automated alternatives in accuracy for both identification and quantification. AVAILABILITY: The R package BQuant is available at: http://www.stat.purdue.edu/~ovitek/BQuant-Web/.

Protein quantification in label-free LC-MS experiments.

The goal of many LC-MS proteomic investigations is to quantify and compare the abundance of proteins in complex biological mixtures. However, the output of an LC-MS experiment is not a list of proteins, but a list of quantified spectral features. To make protein-level conclusions, researchers typically apply ad hoc rules, or take an average of feature abundance to obtain a single protein-level quantity for each sample. We argue that these two approaches are inadequate. We discuss two statistical models, namely, fixed and mixed effects Analysis of Variance (ANOVA), which views individual features as replicate measurements of a protein's abundance, and explicitly account for this redundancy. We demonstrate, using a spike-in and a clinical data set, that the proposed models improve the sensitivity and specificity of testing, improve the accuracy of patient-specific protein quantifications, and are more robust in the presence of missing data.

Integrative specimen information service - a campus-wide resource for tissue banking, experimental data annotation, and analysis services.

We present the architecture and approach of an evolving campus-wide information service for tissues with clinical and data annotations to be used and contributed to by clinical researchers across the campus. The services provided include specimen tracking, long term data storage, and computational analysis services. The project is conceived and sustained by collaboration among researchers on the campus as well as participation in standards organizations and national collaboratives.

Development of high throughput dispersive LC-ion mobility-TOFMS techniques for analysing the human plasma proteome.

A technique that combines ion mobility spectrometry (IMS) with reversed-phase liquid chromatography (LC), collision-induced dissociation (CID) and mass spectrometry (MS) has been developed. The approach is described as a high throughput means of analysing complex mixtures of peptides that arise from enzymatic digestion of protein mixtures. In this approach, peptides are separated by LC and, as they elute from the column, they are introduced into the gas phase and ionised by electrospray ionisation. The beam of ions is accumulated in an ion trap and then the concentrated ion packet is injected into a drift tube where the ions are separated again in the gas phase by IMS, a technique that differentiates ions based on their mobilities through a buffer gas. As ions exit the drift tube, they can be subjected to collisional activation to produce fragments prior to being introduced into a mass spectrometer for detection. The IMS separation can be carried out in only a few milliseconds and offers a number of advantages compared with LC-MS alone. An example of a single 21-minute LC-IMS-(CID)-MS analysis of the human plasma proteome reveals approximately 20,000 parent ions and approximately 600,000 fragment ions and evidence for 227 unique protein assignments.

In vivo selection of human hematopoietic cells in a xenograft model using combined pharmacologic and genetic manipulations.

Strategies that increase the ability of human hematopoietic stem and progenitor cells to repair alkylator-induced DNA damage may prevent the severe hematopoietic toxicity in patients with cancer undergoing high-dose alkylator therapy. In the context of genetic diseases, this approach may allow for selection of small numbers of cells that would not otherwise have a favorable growth advantage. No studies have tested this approach in vivo using human hematopoietic stem and progenitor cells. Human CD34(+) cells were transduced with a bicistronic oncoretrovirus vector that coexpresses a mutant form of O(6)-methylguanine DNA methyltransferase (MGMT(P140K)) and the enhanced green fluorescent protein (EGFP) and transplanted into nonobese diabetic/severe combined immunodeficient (NOD/SCID) mice. Mice were either not treated or treated with O(6)-benzylguanine (6BG) and 1,3-bis(2-chloroethyl)-1-nitrosourea (BCNU). At 8-weeks postinjection, a 2- to 8-fold increase in the percentage of human CD45(+)EGFP(+) cells in 6BG/BCNU-treated versus nontreated mice was observed in the bone marrow and was associated with increased MGMT(P140K)-repair activity. Functionally, 6BG/BCNU-treated mice demonstrated multilineage differentiation in vivo, although some skewing in the maturation of myeloid and B cells was observed in mice transplanted with granulocyte-colony stimulating factor (G-CSF)-mobilized peripheral blood compared to umbilical cord blood. Expansion of human cells in 6BG/BCNU-treated mice was observed in the majority of mice previously transplanted with transduced umbilical cord blood cells. In addition, a significant increase in the number of EGFP(+) progenitor colonies in treated versus nontreated mice were observed in highly engrafted mice indicating that selection and maintenance of human progenitor cells can be accomplished by expression of MGMT(P140K) and treatment with 6BG/BCNU.

Hematoprotection and enrichment of transduced cells in vivo after gene transfer of MGMT(P140K) into hematopoietic stem cells.

The overexpression of mutant forms of O(6)-methylguanine-DNA-methyltransferase (MGMT), resistant to the MGMT inhibitor O(6)-benzylguanine (BG), protects hematopoietic cells from the toxicity of combined BG plus O(6)-alkylating agent chemotherapy. To evaluate the feasibility of this approach for clinically relevant O(6)-alkylating agents, combined therapy with BG and two chloroethylnitrosourea-type drugs, ACNU or BCNU, or the triazene derivative temozolomide (TMZ) was investigated in a murine bone marrow transplant model allowing transgenic expression of the highly BG-resistant MGMT(P140K) mutant. Whereas 20/20 control animals transplanted with nontransduced cells died of progressive myelosuppression during therapy, nearly all animals transplanted with MGMT(P140K)-transduced cells survived treatment with BG/ACNU (12/15), BG/TMZ (10/10), or BG/BCNU (5/5). In surviving animals, hematological parameters improved during chemotherapy and pretreatment levels were reestablished during or shortly after therapy. All animals showed enrichment of transgenic granulocytes (range: 15- to 101-fold) and lymphocytes (range: 16- to 55-fold) in peripheral blood, bone marrow, and spleen. No significant differences were observed between individual treatment groups. Serial transplants demonstrated protection in secondary recipients and confirmed the transduction of transplantable stem cells. Thus, these data demonstrate efficient protection from hematotoxicity and substantial enrichment of transgenic cells following MGMT(P140K) gene transfer and treatment with different O(6)-alkylating drugs.