CHANGES.rst

Changes for deenurp

0.1.9-dev

• add subcommand 'dereplicate_named'
• add new clustering method to filter_outliers (--cluster-type RobustSingleLinkage)
• update defaults for filter_outliers
• use vsearch 2.0.3
• fix error in fill_lonely raised by node with no parent
• added deenurp orientate_sequences --out_notmatched_taxids
• address edge case in select_references which too few refs representing a species remain after clustering at CLUSTER_THRESHOLD

0.1.8

• deenurp ncbi_extract_genbank will die after 10 query retries

0.1.7

• replace uclust with vsearch (uclust dropped as dependency)
• bootstrap.sh installs python requirements in specified order with --no-deps
• installs vsearch 1.10.2

0.1.3

• bumping pplacer support to v1.1.alpha17
• adding pandas support to versions 0.17.*
• deenurp orientate_sequences takes a training set (data/types.fasta) and aligns sequences reverse complementing when necessary (GH-34)
• travis caching is setup for pip installs (GH-22)
• fixed pep 440 versioning bug when installing and running unittests (GH-31)
• new deenurp select-references --include-clusters --exclude-clusters --exclude-sequences arguments
• new deenurp fill_lonely --include-taxid --exclude-taxid arguments

0.1.2

• deenurp rdp_extract_sequences now updates old tax_ids instead of dropping them (GH-27)
• deenurp rdp_extract_sequences creates new column taxid_classified instead of dropping tax_ids of unclassified sequences (GH-28)
• function tax_of_genbank no longer returns None if string 'uncultured bacterium' is in the organism name (GH-29)
• expanding type strain designations to include more terms (GH-16)

0.1.1

• gb2csv creates a csv of Genbank records and optionally the references (GH-23)
• using new container based Travis build (GH-24)

0.1.0

• filter-outliers can calculate pairwise distances using cmalign, muscle, or vsearch
• filter-outliers can use hierarchical clustering for outlier detection (--strategy cluster)
• output of filter-outliers --detailed-seqinfo includes centroid, distance from centroid, etc
• add dependencies on pandas, scipy, and scikit-learn
• bootstrap script installs muscle and vsearch
• add subcommand deenurp pairwise-distances
• version number defined using git describe --tags --dirty

0.0.3

• deenurp search-sequences learned option --search-threshold to allow more permissive database searches
• Suppress FastTree messages in the absence of an error

0.0.2

• require Infernal 1.1
• update 16S alignment model (from https://github.com/rdpstaff/fungene_pipeline/blob/eab8ab3751da687b4d6dbd553f6a1d8261d98385/resources/RRNA_16S_BACTERIA/model.cm)
• add bin/boostrap.sh to create execution environment with installed dependencies
• more fine-grained control over threading/multiprocessing