Retreat Spring 2016 - OpenMS/OpenMS GitHub Wiki

Developer meeting (retreat) in Northumbria, UK, 11-18 March 2016

List of topics to discuss:

  • Missing functionality:
    • How to create a complete PRIDE submission with OpenMS (using complete mzIdentML) (issue: #1655)
    • How to generate a complete mzTab with quantitative results from ProteinQuantifier (issue: #1443)
    • new data structures for protein / compound identification and quantification (related to the two points above: complete mzid and complete mzTab, also see https://github.com/OpenMS/OpenMS/pull/1659 for a proposal for small molecule data we need to track)
    • Quantitative workflows with experimental design
    • Proper mass recalibration (MS1 and MS2) and estimation of calibration function (see https://drive.google.com/file/d/0B0Rm__Yp47TEeFhLS3JZRTZxd1E/view?usp=sharing for a previous diploma thesis on this topic)
  • Migration issues:
    • Evaluate how well idXML can be replaced with mzid as the main format for ID data in OpenMS (related issues: #1671, problems: we need to support small molecules, awkward format if created by hand)
    • Migrate OpenMS to C++11/14/17 (issue: #980)
    • Migrate OpenMS to Qt5
    • Move to SeqAn 2.1 (issue: #1313; see also #1824)
    • tool for the conversion of TOPPAS and KNIME workflows (issue: #1865)
  • Building and testing:
    • New workflow tests? (issue: #1827)
    • Define build matrix (e.g. compilers, platforms, when [PR, nightly, release] to run which tests [WF, coverage, packaging] etc.) for new Jenkins builds
    • Automated docker images (e.g. directly via Jenkins)?
  • Releases:
    • Modularization of OpenMS
    • Semantic versioning
    • Tracking of de.NBI user statistics
  • Other topics:
    • What happened to the proposed changes to the kernel discussed at the Izmir meeting (e.g. represent spectra/chromatograms as two double vectors rather than vectors of 2D points)?
    • AASequence/ResidueDB/ModificationsDB/ResidueModification performance and thread safety issues
    • Tighter semantics integration (CV in enzymesdb style replacement for customary metadata tags, related to workflows with experimental design and #1716,#1409,#1468,...)
    • new QC terms for PTXQC (Chris and Mathias)
    • Combining identifications from different search engines (issue: #1823)
    • Migation from sourceforge - what to do with information in https://sourceforge.net/p/open-ms/wiki/Home/

Schedule:

We propose a relatively fixed schedule of discussion and coding sessions. The details will be filled in as we go along based on above list. In general we propose:

  • Breakfast
  • 09:00-12:00 : Work session (except for 2 people for the kitchen team helping from 11 on)
  • 12:00-14:00 : Lunch break
  • 13:00-18:30 : Afternoon session
  • Evening: Social time, drinking, leisure coding, etc. (no commits > 1.0 promille)

Day 1: Saturday 03/12/2016 (kitchen helper: Xiao and Julianus)

  • Morning
    • Who is working on what? Short talks of participants (if applicable):
      • Xiao and Julianus: Protein inference (20 min)
      • Petra: phosphoproteomics (10 min)
      • Hendrik: proteogenomics (10 min)
      • ?
      • ?
  • Afternoon
    • Discussion of all topics and formation of work groups (can be proposed beforehand)
    • WG: peptide indexer and requirements for SeqAn (Reinert, Andreotti, ??)
    • WG: protein inference plans
    • WG homepage
    • Kernel redesign
    • 18:00 Wrap up for the day

Day 2: Sunday 04/12/2016 (kitchen helper: Sandro and Chris)

  • Morning
    • Discussion of all topics and formation of work groups.
    • WG: Phoning home (Reinert, Kohlbacher, Sachsenberg..)
  • Afternoon
    • Discussion of afternoon plans (30 minutes)
    • 18:00 Wrap up for the day

Day 3: Monday 05/12/2016 THIS WAS THE OUTING DAY

Day 4: Tuesday 06/12/2016

  • Morning
    • Discussion of all topics and formation of work groups.
  • Afternoon
    • Discussion of afternoon plans (30 minutes)
    • 18:00 Wrap up for the day

Day 5: Wednesday 07/12/2016 (Matthias and Eugen and Lars)

  • Morning

    • Discussion of all topics and formation of work groups.

  • Result discussion:

    • Fixed Protein coverage computation (Lars)
    • Adapted OpenMS to use latest version of SeqAn (2.1). It worked with C++-11. (Sandro)
    • Sketched internal data formats roadmap (Matthias, made a Google doc), will start implementation
    • Idea for ProteinIndexer was discussed, will formulate BSc thesis content (Knut)
    • Trying to make imputation work (Tjeerd, Isabel, Wengquang)
    • Working on MSGF+ adapter problem
    • C++ inference work (Xiao)
    • Extending the content of the tutorial (in Doxygen) (Petra)

  • Afternoon

    • Discussion of afternoon plans (30 minutes)

    • 18:00 Wrap up for the day

Day 6: Thursday 08/12/2016 (kitchen helper: Yasset and Petra)

  • Morning
    • Discussion of all topics and formation of work groups.
  • Afternoon
    • Discussion of afternoon plans (30 minutes)
    • 18:00 Wrap up for the day

Day 7: Friday 09/12/2016

  • breakfast and vacate premises