LCMS analysis with envipy Uwe Schmitt Scientific IT Services ETH - PowerPoint PPT Presentation

LCMS analysis with envipy Uwe Schmitt Scientific IT Services ETH Zürich | 28.09.16 | 1 Uwe Schmitt

envipy graphical user interface | 28.09.16 | 2 Uwe Schmitt

envipy graphical user interface principles The graphical user interface: § configures and controls workflow execution § observes running workflow § allows interactive inspection of intermediate and final results § allows export of intermediate results to csv | 28.09.16 | 3 Uwe Schmitt

envipy sample inspector | 28.09.16 | 4 Uwe Schmitt

envipy peaklist inspection | 28.09.16 | 5 Uwe Schmitt

envipy workflow manager § Multi core execution § Always runs workflow from the beginning § Caching of intermediate results § Decoupled from user interface § Headless execution on servers § Every step attaches new information to peaks | 28.09.16 | 6 Uwe Schmitt

envipy workflow: peak picking § enviPick: fast peak picker for high resolution data § Handles .raw, .mzML and .mzXML files § Tested for Orbitrap and TOF data § Assigns MS/MS data from data dependent measurements | 28.09.16 | 7 Uwe Schmitt

envipy workflow: isotope pattern computations § enviPat : fast and machine dependent calculation of isotope patterns § Computes patterns of: § Internal standards § Targets § Suspects | 28.09.16 | 8 Uwe Schmitt

envipy workflow: recalibration / alignment § Reuses code from enviMass § m/z recalibration § rt alignment § Requires internal standards | 28.09.16 | 9 Uwe Schmitt

envipy workflow: peak shape fitting § Asymmetric and symmetric peak shape models § Raw area computation | 28.09.16 | 10 Uwe Schmitt

envipy workflow: screening § Screening routine from enviMass § Matches precalculated isotope patterns to picked peaks § Relaxed rt tolerances for suspect screening | 28.09.16 | 11 Uwe Schmitt

envipy workflow: component assignment § Based on nontarget R package § Assigns component id based on isotope, adduct and homologue relations § Plus: m0 computation | 28.09.16 | 12 Uwe Schmitt

envipy workflow: time series analysis § groups peaks over samples § „greedy“ algorithm for speed + reassignment § from enviMass : scoring of interesting time series § Clustering of time series § Event detection in time series | 28.09.16 | 13 Uwe Schmitt

envipy time series clustering § based on affinity propagation algorithm § No prior knowledge about number of expected clusters § Similarity: correlation without mean normalization | 28.09.16 | 14 Uwe Schmitt

envipy workflow: event detection § smoothing + detection of local maxima produce „seeds“ § Fit to gaussian profiles | 28.09.16 | 15 Uwe Schmitt

envipy workflow: MS/MS inspection • The envipy peak picker extracts MS2 spectra from data dependent MS2 scans • MS2 spectra can be exported as well as inspected in the envipy graphical user interface | 28.09.16 | 16 Uwe Schmitt

envipy MS/MS based identification § Selected peaks with assigned MS/MS spectra can be identified by querying http://msbi.ipb-halle.de/MetFrag/ webservice | 28.09.16 | 17 Uwe Schmitt

envipy: availabilty § Installer for Windows (includes R + Python) § Requires 64bit Windows 7 or newer § Faster with SSD drives § Short release cycle, expect updates. § Runs on Linux + Mac OS X (contact me for details) | 28.09.16 | 18 Uwe Schmitt

envipy: technical details § Mostly Python 2.7 § Reuses existing R code § Starts R in separate process § Communication over pipes § Some speed optimizations in C (from Cython) § Many functionalities from emzed (http://emzed.ethz.ch) § Tables partly based on HDF5 § PyQt4 for the graphical user interface | 28.09.16 | 19 Uwe Schmitt

Contact us Heinz Singer (EAWAG) heinz.singer@eawag.ch Uwe Schmitt (ETH Zürich) uwe.schmitt@id.ethz.ch | 28.09.16 | 20 Uwe Schmitt

Thank You