Major central nervous system complications after allogeneic stem cell transplantation: A large retrospective study on 888 consecutive adult patients.

OBJECTIVES: Major complications affecting the central nervous system (CNS) present a challenge after allogeneic stem cell transplantation (allo-SCT). METHODS: Incidence, risk factors, and outcome were retrospectively analyzed in 888 patients in a monocentric study. RESULTS: Cumulative incidence (CI) of major CNS complications at 1 year was 14.8% (95%CI 12.3%-17.2%). Median follow-up is 11 months. CNS complications were documented in 132 patients: in 36 cases, classified metabolic; 26, drug-related neurotoxicity (14 attributed to cyclosporine A, 4 to antilymphocyte globulin); 11, cerebrovascular (ischemic n = 8, bleeding n = 3); 9, infections; 9, psychiatric; and 9, malignant. The cause of CNS symptoms remained unclear for 37 patients (28%). Multivariate analysis demonstrated an association of CNS complication with patient age (P < .001). The estimated OS of patients with any CNS complication was significantly lower than in patients without neurological complications (P < .001), and the CI of non-relapse mortality (NRM) was higher for patients with CNS complication (P < .001). A significant negative impact on survival can only be demonstrated for metabolic CNS complications and CNS infections (NRM, P < .0001 and P = .0003, respectively), and relapse (P < .0001). CONCLUSION: CNS complications after allo-SCT are frequent events with a major contribution to morbidity and mortality. In particular, the situations of unclear neurological complications need to be clarified by intensive research.

Quantitative comparison of a human cancer cell surface proteome between interphase and mitosis.

The cell surface is the cellular compartment responsible for communication with the environment. The interior of mammalian cells undergoes dramatic reorganization when cells enter mitosis. These changes are triggered by activation of the CDK1 kinase and have been studied extensively. In contrast, very little is known of the cell surface changes during cell division. We undertook a quantitative proteomic comparison of cell surface-exposed proteins in human cancer cells that were tightly synchronized in mitosis or interphase. Six hundred and twenty-eight surface and surface-associated proteins in HeLa cells were identified; of these, 27 were significantly enriched at the cell surface in mitosis and 37 in interphase. Using imaging techniques, we confirmed the mitosis-selective cell surface localization of protocadherin PCDH7, a member of a family with anti-adhesive roles in embryos. We show that PCDH7 is required for development of full mitotic rounding pressure at the onset of mitosis. Our analysis provided basic information on how cell cycle progression affects the cell surface. It also provides potential pharmacodynamic biomarkers for anti-mitotic cancer chemotherapy.

Phosphoproteomics.

Current analytical protein methods show phosphorylation to be the most ubiquitous, evolutionary conserved post-translational modification Post-Translational Modification (PTM). The reversible and transient nature of protein phosphorylation allows signal transduction pathways to carry out diverse cellular functions. From bacteria to humans, phosphorylation serves to modify protein function by altering protein stability, cellular location, substrate affinity, complex formation, and activity; thus allowing essential events such as cell cycle and growth to occur at precise times and locations. Phosphorylation controls a variety of events at many biological levels including: housekeeping activities controlled by single cells such as DNA transcription, cell-cycle regulation, and energy metabolism; and cellular processes that involve signaling between cells or the environment including such as neuronal migration and immune system recognition. This review summarizes state-of-the-art proteomics technologies available to study phosphorylation in biological systems. We highlight the tremendous steps the field has made in the last 5 years which allow quantitative global analyses while pointing out caveats in experimentation.

Estimating the confidence of peptide identifications without decoy databases.

Using decoy databases to compute the confidence of peptide identifications has become the standard procedure for mass spectrometry driven proteomics. While decoy databases have numerous advantages, they double the run time and are not applicable to all peptide identification problems such as error-tolerant or de novo searches or the large-scale identification of cross-linked peptides. Instead, we propose a fast, simple and robust mixture modeling approach to estimate the confidence of peptide identifications without the need for decoy database searches, which automatically checks whether its underlying assumptions are fulfilled. This approach is then evaluated on 41 LC/MS data sets of varying complexity and origin. The results are very similar to those of the decoy database strategy at a negligible computational cost. Our approach is applicable not only to standard protein identification workflows, but also to proteomics problems for which meaningful decoy databases cannot be constructed.

Effect of high-accuracy precursor masses on phosphopeptide identification from MS3 spectra.

This contribution investigates the impact of precursor ion accuracy on the number of phosphopeptides identified from neutral loss-triggered MS(3) spectra acquired on high-accuracy instruments. We show that replacing the MS(3) precursor (parent) ion mass by a highly accurate MS(2) precursor (grandparent) mass increases the number of identified phosphopeptides 3- to 5-fold. The number of unique phosphopeptides identified in addition to the MS(2) data increases from 6% to 24%. Our results illustrate that the sensitivity of phosphopeptide identification depends critically on the mass tolerance setting applied in the database search. This study demonstrates that MS(3) parent ion masses should always be substituted by high-accuracy grandparent ion masses and that error tolerance settings should be set as low as instrument accuracy permits to yield the best possible identification sensitivity.

Non-linear classification for on-the-fly fractional mass filtering and targeted precursor fragmentation in mass spectrometry experiments.

MOTIVATION: Mass spectrometry (MS) has become the method of choice for protein/peptide sequence and modification analysis. The technology employs a two-step approach: ionized peptide precursor masses are detected, selected for fragmentation, and the fragment mass spectra are collected for computational analysis. Current precursor selection schemes are based on data- or information-dependent acquisition (DDA/IDA), where fragmentation mass candidates are selected by intensity and are subsequently included in a dynamic exclusion list to avoid constant refragmentation of highly abundant species. DDA/IDA methods do not exploit valuable information that is contained in the fractional mass of high-accuracy precursor mass measurements delivered by current instrumentation. RESULTS: We extend previous contributions that suggest that fractional mass information allows targeted fragmentation of analytes of interest. We introduce a non-linear Random Forest classification and a discrete mapping approach, which can be trained to discriminate among arbitrary fractional mass patterns for an arbitrary number of classes of analytes. These methods can be used to increase fragmentation efficiency for specific subsets of analytes or to select suitable fragmentation technologies on-the-fly. We show that theoretical generalization error estimates transfer into practical application, and that their quality depends on the accuracy of prior distribution estimate of the analyte classes. The methods are applied to two real-world proteomics datasets. AVAILABILITY: All software used in this study is available from http://software.steenlab.org/fmf CONTACT: hanno.steen@childrens.harvard.edu SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Peak intensity prediction in MALDI-TOF mass spectrometry: a machine learning study to support quantitative proteomics.

BACKGROUND: Mass spectrometry is a key technique in proteomics and can be used to analyze complex samples quickly. One key problem with the mass spectrometric analysis of peptides and proteins, however, is the fact that absolute quantification is severely hampered by the unclear relationship between the observed peak intensity and the peptide concentration in the sample. While there are numerous approaches to circumvent this problem experimentally (e.g. labeling techniques), reliable prediction of the peak intensities from peptide sequences could provide a peptide-specific correction factor. Thus, it would be a valuable tool towards label-free absolute quantification. RESULTS: In this work we present machine learning techniques for peak intensity prediction for MALDI mass spectra. Features encoding the peptides' physico-chemical properties as well as string-based features were extracted. A feature subset was obtained from multiple forward feature selections on the extracted features. Based on these features, two advanced machine learning methods (support vector regression and local linear maps) are shown to yield good results for this problem (Pearson correlation of 0.68 in a ten-fold cross validation). CONCLUSION: The techniques presented here are a useful first step going beyond the binary prediction of proteotypic peptides towards a more quantitative prediction of peak intensities. These predictions in turn will turn out to be beneficial for mass spectrometry-based quantitative proteomics.

Evaluation of radiological features for breast tumour classification in clinical screening with machine learning methods.

OBJECTIVE: In this work, methods utilizing supervised and unsupervised machine learning are applied to analyze radiologically derived morphological and calculated kinetic tumour features. The features are extracted from dynamic contrast enhanced magnetic resonance imaging (DCE-MRI) time-course data. MATERIAL: The DCE-MRI data of the female breast are obtained within the UK Multicenter Breast Screening Study. The group of patients imaged in this study is selected on the basis of an increased genetic risk for developing breast cancer. METHODS: The k-means clustering and self-organizing maps (SOM) are applied to analyze the signal structure in terms of visualization. We employ k-nearest neighbor classifiers (k-nn), support vector machines (SVM) and decision trees (DT) to classify features using a computer aided diagnosis (CAD) approach. RESULTS: Regarding the unsupervised techniques, clustering according to features indicating benign and malignant characteristics is observed to a limited extend. The supervised approaches classified the data with 74% accuracy (DT) and providing an area under the receiver-operator-characteristics (ROC) curve (AUC) of 0.88 (SVM). CONCLUSION: It was found that contour and wash-out type (WOT) features determined by the radiologists lead to the best SVM classification results. Although a fast signal uptake in early time-point measurements is an important feature for malignant/benign classification of tumours, our results indicate that the wash-out characteristics might be considered as important.