T1K v1.0.6

• Add the option "--partial-intron-noseq" to the t1k-build script to handle IPD-KIR v2.13.0
• Add the options to control post-analysis, .e.g. --post-varMaxGroup to control the SNP calling complexity (#27 )
• Add the option "--gtf-gene-name-mapping" to the AddGeneCoord script to allow the conversion of GTF gene name to the IPD gene names, e.g. HFE:HLA-HFE (#30 )
• Use log-llikelihood difference in the t1k-copynumber script (#28)
• Fix a minor issue for stabling the output from t1k-build
• The "--od" option can support recursive folder creation

T1K v1.0.5

• Fix a serious bug that put the imputed intron off by 1.
• Output an extra column in the genotype.tsv file for the filtered but with highly abundance alleles. This information might be useful for copy number inference in future versions.

T1K v1.0.4

When extracting the reference sequences from the IPD-HLA/KIR database, the new script will keep the partial alleles (partial: either explicitly marked as partial or have missing intron) that have complete exon sequences. The missing intron anchors for those partial alleles will be filled by the most commonly used intron anchor sequences from the same gene.

This may hugely impact the KIR allele predictions as many prevalent alleles, such as KIR3DL3*005, have missing introns but complete exon structures in the IPD-KIR database.

T1K v1.0.3

• Change the link to t1k.dat file on IPD-KIR, as the 2.12.0 is moved to kir.dat file instead of the KIR.data (stores 2.11.0) on their website
• for the RNA reference file, keep the alleles marked partial but complete in RNA level in IPD dat file. This is because IPD-KIR 2.11.0 marks many partial DNA sequences explicitly, and seems to remove them in IPD-KIR 2.12.0. In case IPD-KIR annotates partial DNAs again, we add this behavior when extracting alleles.
• Add the option "--outputReadAssignment" to show the read-to-allele assignment details. (#12 )
• Update the README to clarify the output format. (#15 )

T1K v1.0.2

• Add the option "--alleleWhitelist" to only retain alignments on the alleles in the whitelist. An example of creating the whitelist based on OptiType's frequent allele list can be found at "https://github.com/mourisl/T1K_manuscript_evaluation/tree/master/Fig1"
• Tuned the parameter in the SMART-seq wrapper
• Adjust the weight in the EM algorithm if the allele length differs too much.

T1K v1.0.1

• Fix several bugs that may cause crashes.

T1K v1.0.0

The Initial public release.