MQLS ("More Poweful" or "Modified" Quasi-likelihood Score Test) Changelog Version 1.1 (8/01/07) ======================= - Initial Release Version 1.2 (3/20/08) ======================= Daniel E. Weeks (weeks@pitt.edu) -- Corrected memory allocation so that MQLS should not as easily run out of memory when analyzing datasets with a large number of people. -- Improved the handling of files with too many markers: It now prints an error message and exits if the maximum number of markers (MAXMARK) has not been set large enough. Temporary fix (6/16/08) ========= Mary Sara McPeek -- In MQLS.c, changed printing of p-values so that they will be in either scientific or floating point notation, whichever is shorter. This is intended to eliminate the problem of very small p-values being reported as .000000 in floating point notation. Have not updated the executable MQLStest. Version 1.3 (2/23/09) ==================== Mary Sara McPeek a) Fixed another memory allocation problem. b) Changed printing of p-values to sigfile (in addition to outfile) so that they will be in either scientific or floating point notation, whichever is shorter. c) Changed the way MZ twins are identified. Version 1.4 (9/23/09) ==================== Timothy Thornton a) Fixed a problem that previouly crashed the program when there are no members of a family with available genotyped data at a SNP. Mark Kvale (kvalem@humgen.ucsf.edu) a) Fixed memory allocation problem where family struct is now allocated only once for each family instead of for every marker b) Fixed memory allocation problem where memory is now deallocated for the temporary matrices in the makeFreqMat subroutine Version 1.5 (3/14/11) ==================== Timothy Thornton a) All three association statistics are now calculated using the variance estimator of Equation (3) of Thornton and McPeek (2010), which is a robust variance estimator that relaxes the Hardy-Weinberg Equilibrium (HWE) assumption under the null hypothesis. The previous versions of the MQLS software were calculated using a variance estimator that assumes HWE in the founders. b) Modified the software to print the p-values for every SNP to a file named "MQLStest.pvalues" c) Changed the way MZ twins are identified by also confirming that pairwise kinship coefficient value for twins equals the self-kinship coefficient values. Also corrected an error in the MZ twin subroutine.