HIBLUP Downloads

HIBLUP is written in C++, the released executable files can be used in terminal of different platform directly, there is no process of installation.
Binary file for different platform can be found below:

After downloaded successfully, please type ./hiblup --help in the terminal to see all available functions and parameters.
The test dataset and example script prepared for running HIBLUP can be downloaded by clicking demo.zip.
For more details, please refer to the Online User Manual.

Update history

Version 1.5.3 _{28 Sep 2024}

• fixed several bugs:
  error for multiple traits model of unbalanced records in certain situation
  mixed-use of character match function leads to unexpected LD/LDscore output

Version 1.5.0 _{1 Aug 2024}

• add following new functions:
  genotype format conversion (--trans-geno)
  rewrite the module of multiple traits model to support both balanced and unbalanced records
  support ignoring residual covariance when fitting multiple traits model (--ignore-cove)
• fixed several bugs:
  solve MME of single trait model
  computing SNP effect when the flag --r2 is specified

Version 1.4.0 _{2 Feb 2024}

• add following new functions:
  LD score computing (--ldsc)
  LD score regression (--ldsr) to estimate heritability and genetic correlation using summary level data
  multiple traits SBLUP model (MT-SBLUP)
  Mixed GRM construction (--pop-class)
  Prediction using individual genotype and SNP effects (--pred)
  support loading file containing user-defined windows (--window-file)
  support using provided GRM when constructing H matrix for SSBLUP model
• some improvements:
  greatly improved the computational efficiency of SBLUP model
  significance test of estimated heritability and genetic correlation
• fixed several bugs:
  HE regression for multiple traits
  G by E interaction
  marker effect estimation for special cases
• delete the license activation module

Version 1.3.1 _{20 Feb 2023}

• add following new functions:
  LD correlation matrix (--ld)
  summary data BLUP (SBLUP) model (--sblup)
  transform PLINK binary genotype file into BLUPF90 required format (--bf2f90)
  output a readable TXT format file when constructing relationship matrix or LD matrix (--write-txt)
• new parameters: --window-bp, --window-num, --window-geno, --sumstat, --ldm, --h2.
• remove LOG coloring for Windows platform.
• Some software architecture improvements:
  use Docker for multi-platform packaging
  libcurl is converted to static linking, which is used to activate license and support syslog
  license module now supports multiple machines at the same time

Version 1.2.0 _{19 Aug 2022}

• improve the computational efficiency and lower the memory cost for HE regression.
• further reduce the memory cost for GRM construction.
• improve the computational efficiency for PCG algorithm.
• add a parameter (--he-pred) to control wether to compute the random effects when using HE regression.

Version 1.1.0 _{15 Apr 2022}

• add the following functions:
  allele frequency (--allele-freq) and genotype frequency (--geno-freq)
  homozygosity (--homo) and heterozygosity (--hete)
  different algorithms for GRM construction (--code-method)
  inbreeding coefficients (--ibc) and relationship coefficients (--rc) from pedigree or genotype
  reliability and PEV for all random effects (--r2)
  PCA (--pca, --npc)
• fix some bugs and improve the computational efficiency.

Version 1.0.603 _{26 Nov 2021}

• fixed a bug for constructing GRM.
• fixed the problem of output alignment of full-width characters.

Version 1.0.592 _{18 Nov 2021}

• fixed a bug for LU decomposition in variance components estimation.
• added LU decomposition in MME solver.

Version 1.0.589 _{21 Oct 2021}

• fixed some bugs proposed by users for a more inclusive experience.