GitLab

Overduin, Sam authored Feb 24, 2020 and Overduin, Sam committed Feb 24, 2020

Signed-off-by: Overduin, Sam <sam.overduin@wur.nl>

Overduin, Sam authored Feb 21, 2020 and Overduin, Sam committed Feb 21, 2020

config.info:
- Added taxonomy false.

barcode_index_builder.hpp:
- only extract TaxaTree if cfg::get().taxonomy

scaffold_graph_construction_pipeline.cpp:
- only add TaxaBreak constructor if cfg::get().taxonomy

config_struct.cpp & hpp:
- added taxonomy to cfg.

barcode_index_construction.cpp:
- Only decide LCA if cfg::get().taxonomy

spades_iteration_stage.py:
- Add taxonomy option to config.info of final iteration

options_parser.py:
- Add --taxonomy argument. Only usable with --meta and --gemcode options.

Some various other clean-ups in files:
- aa_todo_notes
- path_polisher.hpp
- path_scaffolder.cpp
- scaffold_vertex.cpp
Signed-off-by: Overduin, Sam <sam.overduin@wur.nl>

Overduin, Sam authored Feb 06, 2020 and Overduin, Sam committed Feb 06, 2020

scaffold_vertex.hpp & cpp:
- Used Fuzzy LCA for PathVertex::GetTaxaTree with max_drilldown_ratio=0.4,
  max_mismatch_ratio=0.03 and min_assignment_ratio=0.3. This code is in
  barcode_index_construction.cpp

barcode_index.hpp & barcode_index_builder.hpp:
- added code to deal with "_SUBSTR(0,106)" at end of read names (influenced
 taxatrees).

launcher.cpp:
- Removed LCA parameter testing to start of repeat resolving step

read_cloud_connection_conditions.cpp & hpp:
- Output which edge connections (or path connections) are being broken in
  logfile.

scaffold_graph_construction_pipeline.cpp:
- Moved TaxaBreakConstructor to bottom of pipelines for more informative
  breaks.

path_scaffolder.cpp:
- map.at() replaced with map[] if item not within map. Added debuggin msg.

barcode_index_construction.cpp:
- Added function to get LCA for PathScaffoldVertex (more conservative than
  reads-> contig LCA).
- Started prototype code for taxonomy assignment per edges (eg. using CAT)
Signed-off-by: Overduin, Sam <sam.overduin@wur.nl>

Overduin, Sam authored Jan 15, 2020 and Overduin, Sam committed Jan 15, 2020

scaffold_vertex.hpp & cpp:
- Made GetTaxaTree() member of PathVertex and EdgeIdVertex (both are
  ScaffoldVertexes in the contracted DeBruijnGraph)

launcher.cpp:
- Added LCA parameter testing to start of repeat resolving step

read_cloud_connection_conditions.cpp & hpp:
- Made GetTaxaTree() part of the ScaffoldVertex itself (Edge- & Path-type)

scaffold_graph_construction_pipeline.cpp:
- Moved TaxaBreakConstructor to top of pipelines for efficiency

barcode_index_construction.cpp:
- Turned LCA settings tester back on

barcode_index_construction.cpp & hpp:
- Re-implemented taxonomy assignment with fuzzy_lca
- Some cleaner code in taxonomy assignment
- Added parameter testing module
- Bugfixes, header file updated

read_cloud_connection_conditions.cpp & hpp:
- Code cleanup (remove duplicated logic)

construction_callers & scaffold_graph_construction_pipeline:
- Nothing (just code that was commented out again)
Signed-off-by: Overduin, Sam <sam.overduin@wur.nl>

Overduin, Sam authored Jan 08, 2020 and Overduin, Sam committed Jan 08, 2020

barcode_index.hpp:
- Improved robustness of some functions (error handling)
- Added taxid_distribution serialisation

barcode_index_construction.cpp & hpp:
- Re-implemented taxonomy assignment with fuzzy_lca
- Some cleaner code in taxonomy assignment
- Added parameter testing module
- Bugfixes, header file updated

read_cloud_connection_conditions.cpp & hpp:
- Code cleanup (remove duplicated logic)

construction_callers & scaffold_graph_construction_pipeline:
- Nothing (just code that was commented out again)
Signed-off-by: Overduin, Sam <sam.overduin@wur.nl>

Overduin, Sam authored Dec 06, 2019 and Overduin, Sam committed Dec 06, 2019

barcode_index.hpp:
- Implemented taxa_codes + taxonomy_ serialisation
- Made taxa_codes::map's public

barcode_info_extractor.hpp:
- Added funcs: EdgeIsValid, GetConjugateEdge, GetSequence, GetEdgeLength

barcode_index_construction.cpp:
- Added write_contigs_to_fasta for testing purposes.
- Re-implemented taxonomy assignment to use edge & conjugate.

read_cloud_connection_conditions.cpp:
- In TaxaBreakPred.: if taxonomy of edge not found, try taxonomy of
  conjugate edge. Else taxa 0.

logger.hpp:
- Created special DEBUG_SAM for personal logging. Will change at end.

Overduin, Sam authored Dec 03, 2019 and Overduin, Sam committed Dec 03, 2019

barcode_index.hpp:
- Initialized TaxIdEncoder with "0" taxatree
- Added func TaxIdEncoder::ToTaxaTreeVector
- Bugfix EdgeEntry::GetTaxonomy()

barcode_info_extractor.hpp:
- Added funcs: ToTaxaTreeVector, TaxaTreeFromTaxId, GetTaxaTreeFromEdge

barcode_index_construction.cpp:
- Moved ToTaxaTreeVector to extractor & barcode_index
- Improved majority_vote_lca to actually use LCA

construction_callers.cpp & hpp:
- Added TaxaBreakConstructorCaller to break mismatched taxa in
  contracted assembly graph

read_cloud_connection_conditions.cpp:
- Added TaxaBreakPredicate that returns false when transition
  (scaffold_edge) has incompatible taxonomy

scaffold_graph_construction_pipeline.cpp:
- Added TaxaBreak routine to basic mode & scaffolding mode

Overduin, Sam authored Nov 29, 2019 and Overduin, Sam committed Nov 29, 2019

- Added no TaxaTree in read means taxa '0'
- Bugfix taxatree_prefixes
- Cut out taxatree parser to vector<int>

barcode_index.hpp:
- Changed TaxIdEncoder to fully use TaxId vs int64
- small bugfixes eg: moved declarations around, corrected taxid_begin()
- made edge_to_entry in FrameBarcodeIndex public

barcode_info_extractor.hpp:
- Added GetTaxidCount

barcode_index_construction.cpp:
- Added chunk of code to determine and set LCA per EdgeEntry in barcode_index

Overduin, Sam authored Nov 27, 2019 and Overduin, Sam committed Nov 27, 2019

- Added TaxaTree (like 1.23.456.7890) parser from reads & store in index

barcode_index.hpp:
- Added data structure in BarcodeIndex to store TaxaTree mapping
  to taxids (taxatree_codes_).
- Added data structure in EdgeEntry to store taxid_distribution_.
- Added attribute to EdgeEntry for taxonomy_.

barcode_info_extractor.hpp:
- Added method to get taxids for a EdgeID

Changed scope of static const MAGIC_LOOP_CONSTANT due to CLion build process complaining. Shouldn't break anything.
Also uploaded aa_todo_notes. In future will not mention this file.