adVNTR-1.5.0

• Add duplicated read check
• Log read ID and read source (mapped/unmapped)
• [PacBio] Add VCF output option for PacBio mode
• [PacBio] Add a flag to log Pacbio reads
• Performance improvement (calling viterbi algorithm)
• [pairwise_alignment generator] support read_id and read_source (mapped/unmapped)
• Bug fixes on VNTR DB manipulation
• Minor bug fixes

adVNTR-1.4.1

• Print all class of reads in log file
• Fix HMM threshold to discard aligned reads
• Pairwise alignment–shows flanking reads as well
• Pairwise alignment–aligns each repeat unit with the most similar repeat unit in the reference
• Minor bug fixes

adVNTR-1.4.0

• Add VCF output format
• Add Neural Network filtering in the master/main branch
• New pairwise-alignment tool for verifying genotypes
• Reference VNTR models for GRCh38
• More stringent filtering

adVNTR-1.3.3

• Fix bug in addmodel with custom directories
• Disable multiprocessing in read recruitment as python multiprocessing sometimes crashes
• Improvements in HMM for genotyping

adVNTR-1.3.2

• Fix memory issue in filtering

adVNTR-1.3.1

• Improve keyword matching for non-unique keywords
• Accept remote and custom reference for CRAM files
• Bug fixes, mainly for specific VNTR in INS gene

adVNTR-1.3.0

• Add support for CRAM files
• Genotype all VNTRs in database and provide different pre-trained databases
• Reduce memory usage of addmodel command
• Fix some problems in Python 3.6.5

adVNTR-1.2.0

• Achieve 2X speedup since the last version
• New C++ module to do the keyword matching in a more efficient way
• Consider more comprehensive annotation for VNTRs and include more VNTRs in the genotyping
• Add the option to update the models on a sample
• Add BED output format

adVNTR-1.1.1

• Add --haploid parameter for haploid organisms
• Improve PacBio models
• Doesn't necessarily require hg19_VNTRs.db to run

adVNTR-1.1.0

• New addmodel command to add and train new VNTRs.
• Use LinearRegression to train models. It finds a threshold for HMM scores to recruit reads.
• Save all information of models in a unified database and remove other files