syntax for calling macro and sample output

/* HEADING */

/* This program is used to test WilGoonessFit.sas macro */

filename crd '/udd/nhron/macrodevelopment/WilcxRank/data/IlluminaSQNM.prn';

libname inflb '/udd/nhron/macrodevelopment/WilcxRank/data';

OPTIONS PAGESIZE = 60 LINESIZE = 100 PAGENO = 1 date formdlim=']';

*options symbolgen mprint;

options yearcutoff=1920;

footnote '/udd/nhron/macrodevelopment/WilcxRank/testWilgoodnessfit.sas';

proc format;

value ag 1='< 60' 2='60-69' 3='70-79' 4='80+';

value ag1f 1='< 60-69' 2='70-95' ;

value bm 1='<25.0' 2='25.0-29.9' 3='30.0+';

value sx 1='Male' 2='Female';

run;

data sqnm;

infile crd firstobs=2;

input sbj_uid 2-11 @73 rs1061170 $char3.;

run;

proc sort; by sbj_uid;

data ids;

set inflb.illuminastudyids end=eof;

retain fam twn prog regg areds 0;

if fhid ne ' ' then fam=fam+1;

if meeiid ne ' ' then areds=areds+1;

if pstudyidwop ne ' ' then prog=prog+1;

if rstudyidwor ne ' ' then regg=regg+1;

if sbj_twn_caseid ne ' ' then twn=twn+1;

if eof then put fam= areds= prog= regg= twn= ;

run;

proc sort; by sbj_uid;

data amdgrd;

set inflb.origgradelist;

run;

proc sort; by sbj_uid;

data demog;

set inflb.jsdb_sbj_subjects;

run;

proc sort; by sbj_uid;

data all;

merge ids(in=ia) amdgrd(in=ib) demog(in=ic) sqnm(in=ie); by sbj_uid;

if ia and ib and ic and ie;

run;

proc sort; by pstudyidwop;

/* reading the genotyping data */

proc import out=aredsphenotype1

datafile="/udd/nhron/macrodevelopment/WilcxRank/data/AREDS_phenotype_072605.csv"

dbms=dlm replace;

delimiter=',';

getnames=yes;

run;

proc import out=aredsphenotype2

datafile="/udd/nhron/macrodevelopment/WilcxRank/data/AREDS_phenotype_072905.csv"

dbms=dlm replace;

delimiter=',';

getnames=yes;

run;

data aredsphenotype2(drop=Baseline_Age Age_at_Last_Photos);

set aredsphenotype2;

where Obs^=.;

f1=input(Baseline_Age,8.);

f2=input(Age_at_Last_Photos,8.);

run;

data aredsphenotype2;

set aredsphenotype2;

rename f1=Baseline_Age

f2=Age_at_Last_Photos;

run;

proc sort data=aredsphenotype1;

by specimen_number;

run;

proc sort data=aredsphenotype2;

by specimen_number;

run;

data mergeemmes;

set aredsphenotype1 aredsphenotype2;

by specimen_number;

run;

proc import out=newemmesgrades

datafile="/udd/nhron/macrodevelopment/WilcxRank/data/AREDS_EMMES_revised_phenotype_20051122.csv"

dbms=dlm replace;

delimiter=',';

getnames=yes;

run;

data newemmesgrades(drop=obs

rename=(specimen_number=specimen_number_new));

set newemmesgrades;

run;

proc sql;

create table newmergeemmes as select mergeemmes.*, newemmesgrades.*

from mergeemmes left join newemmesgrades on

mergeemmes.specimen_number=newemmesgrades.specimen_number_new;

run;

proc import out=sqnm

datafile="/udd/nhron/macrodevelopment/WilcxRank/data/Broad_Illumina_SQNM_results_20050728.csv"

dbms=dlm replace;

delimiter=',';

getnames=yes;

run;

proc sort data=sqnm;

by individual_id;

run;

proc sql;

create table emmesaredstemp as select sqnm.*, newmergeemmes.* from

sqnm left join newmergeemmes on sqnm.individual_id=newmergeemmes.specimen_number;

run;

data emmesareds;

set emmesaredstemp;

where specimen_number^="";

/* note: missing some last photo dates because of formatting issue above */

*mostrecent_grade=current_amd_category;

smokstat=baseline_smoking_status;

age_recent=Current_Age;

if smokstat='Former' then smokstat='PAST';

if smokstat='Current' then smokstat='CURRENT';

if smokstat='Never' then smokstat='NEVER';

if sex='M' then ssex=1;

if sex='F' then ssex=2;

if . < baseline_bmi < 25.0 then bmigp=1;

else if 25.0 <= baseline_bmi < 30.0 then bmigp=2;

else if 30.0 <= baseline_bmi then bmigp=3;

format ethnic $20.;

ethnic=race;

drop sex race;

run;

/* EMMES AREDS data */

data emmesareds;

set emmesareds;

/* calculate education variables */

/* education: lowest common denominator of text education categories */

format education2 $60.;

education2=education;

if education="Some College or Associate Degree"

then education2="Some College or Associate Degree or Vocational Training";

if education="Bachelor's Degree" then education2="College Degree";

if education="Postgraduate Work" then education2="Graduate Work";

/* highsch: 0=less than high school, 1=high school or more */

if education2="Grade 11 or Less" then highsch=0;

if education2 in ("High School Graduate","Some College or Associate Degree or Vocational Training",

"College Degree","Graduate Work") then highsch=1;

/* highsch2: 0=high school or less, 1=more than high school */

if education2 in ("Grade 11 or Less","High School Graduate") then highsch2=0;

if education2 in ("Some College or Associate Degree or Vocational Training",

"College Degree","Graduate Work") then highsch2=1;

/* educatnum: numerical assignation to all all text categories */

format educatnum 3.;

if education="Grade 11 or Less" then educatnum=11;

if education="High School Graduate" then educatnum=12;

if education="Some College or Associate Degree" then educatnum=14;

if education="Bachelor's Degree" then educatnum=16;

if education="Postgraduate Work" then educatnum=17;

emmeseducation=education;

rename education2=education Age_at_Last_Photos=Age_at_Last_Visit ssex=gender;

label Age_at_Last_Photos="Age at Last Visit" ssex=gender;

drop education;

run;

data emmesareds2;

set emmesareds;

mostrecent_grade=MEEI_Current_AMD_category;

Age_at_Last_Photos=Age_at_Last_Visit;

if . < baseline_bmi < 25.0 then base_bmigp=1;

else if 25.0 <= baseline_bmi < 30.0 then base_bmigp=2;

else if 30.0 <= baseline_bmi then base_bmigp=3;

if . < Current_bmi < 25.0 then current_bmigp=1;

else if 25.0 <= Current_bmi < 30.0 then current_bmigp=2;

else if 30.0 <= Current_bmi then current_bmigp=3;

if . < Baseline_Age < 70 then base_agegrp=1;

else if 70 <= Baseline_Age < 95 then base_agegrp=2;

if . < Age_at_Last_Photos < 60 then photo_agegrp=1;

else if 60 <= Age_at_Last_Photos < 70 then photo_agegrp=2;

else if 70 <= Age_at_Last_Photos < 80 then photo_agegrp=3;

else if 80 <= Age_at_Last_Photos < 95 then photo_agegrp=4;

if . < age_recent < 60 then current_agegrp=1;

else if 60 <= age_recent < 70 then current_agegrp=2;

else if 70 <= age_recent < 80 then current_agegrp=3;

else if 80 <= age_recent < 95 then current_agegrp=4;

if rs1061170=0 then delete;

if mostrecent_grade=2 or mostrecent_grade=3 then delete;

if gender='2' then gender=0;

if base_bmigp=1 then bmilt25=1; else bmilt25=0;

if base_bmigp=2 then bmi2529=1; else bmi2529=0;

if base_bmigp=3 then bmi30=1; else bmi30=0;

if base_bmigp=2 or base_bmigp=3 then bmi25=1; else bmi25=0;

if base_bmigp=1 or base_bmigp=2 then bmilt30=1; else bmilt30=0;

if base_agegrp=2 then bage7079=1; else bage7079=0;

if rs1061170=44 then genotp=0;

else if rs1061170=24 then genotp=1;

else if rs1061170=22 then genotp=2;

if genotp=1 then genotp24=1; else genotp24=0;

if genotp=2 then genotp22=1; else genotp22=0;

if genotp=0 then genotp44=1; else genotp44=0;

if smokstat='CURRENT' then smkcur=1; else smkcur=0;

if smokstat='PAST' then smkpst=1; else smkpst=0;

if mostrecent_grade=1 then cascnt=0;

else if mostrecent_grade=4 or mostrecent_grade=5 then cascnt=1;

if Took_Centrum_During_Study_='N' then centrum=0;

else if Took_Centrum_During_Study_='Y' then centrum=1;

if AREDS_Treatment='Zinc' then zinc=1; else zinc=0;

if AREDS_Treatment='Antiox+Zinc' then antzinc=1; else antzinc=0;

if AREDS_Treatment='Antioxidants' then antiox=1; else antiox=0;

if AREDS_Treatment='Antiox+Zinc' or AREDS_Treatment='Antioxidants'

then antioxidant=1; else antioxidant=0;

if smkcur=1 or smkpst=1 then smkevr=1;

if smkcur=0 and smkpst=0 then smkevr=0;

if smkcur=0 and smkpst=0 then smknvr=1; else smknvr=0;

age_geno24=bage7079*genotp24;

age_geno22=bage7079*genotp22;

age_geno44=bage7079*genotp44;

age_genotp=bage7079*genotp;

sex_geno24=gender*genotp24;

sex_geno22=gender*genotp22;

sex_geno44=gender*genotp44;

sex_genotp=gender*genotp;

hisch_geno24=highsch*genotp24;

hisch_geno22=highsch*genotp22;

hisch_geno44=highsch*genotp44;

hisch_genotp=highsch*genotp;

bmilt25_geno24=bmilt25*genotp24;

bmilt25_geno22=bmilt25*genotp22;

bmilt25_geno44=bmilt25*genotp44;

bmilt30_geno24=bmilt30*genotp24;

bmilt30_geno22=bmilt30*genotp22;

bmilt30_geno44=bmilt30*genotp44;

bmi2529_geno24=bmi2529*genotp24;

bmi2529_geno22=bmi2529*genotp22;

bmi2529_geno44=bmi2529*genotp44;

bmi2529_genotp=bmi2529*genotp;

bmi30_geno24=bmi30*genotp24;

bmi30_geno22=bmi30*genotp22;

bmi30_geno44=bmi30*genotp44;

bmi30_genotp=bmi30*genotp;

bmi25_geno24=bmi25*genotp24;

bmi25_geno22=bmi25*genotp22;

bmi25_geno44=bmi25*genotp44;

bmi25_genotp=bmi25*genotp;

smkcur_geno24=smkcur*genotp24;

smkcur_geno22=smkcur*genotp22;

smkcur_genotp=smkcur*genotp;

smkpst_geno24=smkpst*genotp24;

smkpst_geno22=smkpst*genotp22;

smkpst_genotp=smkpst*genotp;

smknvr_geno24=smknvr*genotp24;

smknvr_geno22=smknvr*genotp22;

smknvr_geno44=smknvr*genotp44;

smkevr_geno24=smkevr*genotp24;

smkevr_geno22=smkevr*genotp22;

smkevr_geno44=smkevr*genotp44;

smkevr_genotp=smkevr*genotp;

if ethnic='White';

run;

proc format;

value ag 1='50-59' 2='60-69' 3='70-79' 4='80-95';

value bmi 1='<25.0' 2='25.0-29.9' 3='30.0+';

run;

title 'Grade 4 & 5 vs. Controls';

*proc logistic descending;

*model cascnt=bage7079 gender highsch genotp24 genotp22;

run;

*proc logistic descending;

*model cascnt=bage7079 gender highsch genotp24 genotp22

bmi2529 bmi30 smkcur smkpst ;

run;

data model1;

set emmesareds2;

if cascnt=1 then events=1;

else if cascnt=0 then events=0;

/* From logistic regression above */

a=0.6104;

b1=1.0911;

b2=-0.2749;

b3=-0.8444;

b6=0.2538;

b7=0.6764;

b8=1.4599;

b9=0.4212;

score1=a+b1*bage7079+b2*gender+b3*highsch

+b6*bmi2529+b7*bmi30+b8*smkcur+b9*smkpst;

a=-0.5265;

b1=1.1260;

b2=-0.3102;

b3=-0.6727;

b4=0.9770;

b5=2.0078;

b6=0.1261;

b7=0.7190;

b8=1.6256;

b9=0.5023;

score2=a+b1*bage7079+b2*gender+b3*highsch+b4*genotp24

+b5*genotp22+b6*bmi2529+b7*bmi30+b8*smkcur+b9*smkpst;

run;

data model2;

set emmesareds2;

if cascnt=1 then events=1;

else if cascnt=0 then events=0;

a=0.2270;

b1=1.0501;

b2=-0.1966;

b3=-0.7446;

b4=0.8174;

b5=1.8914;

score1=a+b1*bage7079+b2*gender+b3*highsch+b4*genotp24+b5*genotp22;

a=-0.5265;

b1=1.1260;

b2=-0.3102;

b3=-0.6727;

b4=0.9770;

b5=2.0078;

b6=0.1261;

b7=0.7190;

b8=1.6256;

b9=0.5023;

score2=a+b1*bage7079+b2*gender+b3*highsch+b4*genotp24

+b5*genotp22+b6*bmi2529+b7*bmi30+b8*smkcur+b9*smkpst;

run;

%include "WilGoodnessFit.sas";

%WilGoodnessFit(maindsn=model1, outcome=events, subgrp=base_agegrp,id=individual_id);

Output:

Grade 4 & 5 vs. Controls 10:12 Monday, March 3, 2008 2

------------------------------------------ base_agegrp=1 -------------------------------------------

The CORR Procedure

2 Variables: probit_a probit_b

Simple Statistics

Variable N Mean Std Dev Sum Minimum Maximum

probit_a 285 0.17417 0.97058 49.63936 -2.19015 2.87243

probit_b 285 0.31192 0.93992 88.89717 -1.83027 2.74182

Pearson Correlation Coefficients, N = 285

Prob > |r| under H0: Rho=0

probit_a probit_b

probit_a 1.00000 0.64005

<.0001

probit_b 0.64005 1.00000

<.0001

/udd/nhron/macrodevelopment/WilcxRank/testWilgoodnessfit.sas

]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]

Grade 4 & 5 vs. Controls 10:12 Monday, March 3, 2008 3

------------------------------------------ base_agegrp=2 -------------------------------------------

The CORR Procedure

2 Variables: probit_a probit_b

Simple Statistics

Variable N Mean Std Dev Sum Minimum Maximum

probit_a 289 0.11795 0.96609 34.08679 -1.99014 2.77741

probit_b 289 0.18327 0.92935 52.96450 -2.54401 2.77741

Pearson Correlation Coefficients, N = 289

Prob > |r| under H0: Rho=0

probit_a probit_b

probit_a 1.00000 0.56860

<.0001

probit_b 0.56860 1.00000

<.0001

/udd/nhron/macrodevelopment/WilcxRank/testWilgoodnessfit.sas

]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]

Grade 4 & 5 vs. Controls 10:12 Monday, March 3, 2008 4

------------------------------------------ base_agegrp=1 -------------------------------------------

The CORR Procedure

2 Variables: probit_a probit_b

Simple Statistics

Variable N Mean Std Dev Sum Minimum Maximum

probit_a 205 -0.23374 0.94006 -47.91669 -2.19015 2.28346

probit_b 205 -0.43153 0.88212 -88.46325 -2.45073 2.45073

Pearson Correlation Coefficients, N = 205

Prob > |r| under H0: Rho=0

probit_a probit_b

probit_a 1.00000 0.50185

<.0001

probit_b 0.50185 1.00000

<.0001

/udd/nhron/macrodevelopment/WilcxRank/testWilgoodnessfit.sas

]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]

Grade 4 & 5 vs. Controls 10:12 Monday, March 3, 2008 5

------------------------------------------ base_agegrp=2 -------------------------------------------

The CORR Procedure

2 Variables: probit_a probit_b

Simple Statistics

Variable N Mean Std Dev Sum Minimum Maximum

probit_a 75 -0.43326 0.87768 -32.49484 -1.99014 1.29732

probit_b 75 -0.70387 0.87621 -52.79046 -2.54401 1.38151

Pearson Correlation Coefficients, N = 75

Prob > |r| under H0: Rho=0

probit_a probit_b

probit_a 1.00000 0.50165

<.0001

probit_b 0.50165 1.00000

<.0001

/udd/nhron/macrodevelopment/WilcxRank/testWilgoodnessfit.sas

]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]

Both models 10:12 Monday, March 3, 2008 6

Defference

base_ # of # of Theta +/- se Theta +/- se of Theta from

Obs agegrp Cases Controls for model1 for model2 2 models

1 1 285 205 0.607+/-0.026 0.725+/-0.023 -0.118+/-0.023

2 2 289 75 0.657+/-0.035 0.759+/-0.031 -0.102+/-0.033

/udd/nhron/macrodevelopment/WilcxRank/testWilgoodnessfit.sas

]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]

Both models 10:12 Monday, March 3, 2008 7

# of # of Theta +/- se Theta +/- se Defference of

Cases Controls for model1 for model2 Theta from 2

Obs Summary Summary Summary Summary models Summary pvalue

1 574 280 0.624+/- 0.021 0.737+/- 0.018 -0.113+/- 0.019 2.2186E-9

/udd/nhron/macrodevelopment/WilcxRank/testWilgoodnessfit.sas

]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]

Both models 10:12 Monday, March 3, 2008 9

------------------------------------------ base_agegrp=1 -------------------------------------------

The CORR Procedure

2 Variables: probit_a probit_b

Simple Statistics

Variable N Mean Std Dev Sum Minimum Maximum

probit_a 285 0.26467 0.91999 75.43050 -1.58920 2.35888

probit_b 285 0.31192 0.93992 88.89717 -1.83027 2.74182

Pearson Correlation Coefficients, N = 285

Prob > |r| under H0: Rho=0

probit_a probit_b

probit_a 1.00000 0.78802

<.0001

probit_b 0.78802 1.00000

<.0001

/udd/nhron/macrodevelopment/WilcxRank/testWilgoodnessfit.sas

]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]

Both models 10:12 Monday, March 3, 2008 10

------------------------------------------ base_agegrp=2 -------------------------------------------

The CORR Procedure

2 Variables: probit_a probit_b

Simple Statistics

Variable N Mean Std Dev Sum Minimum Maximum

probit_a 289 0.13732 0.92075 39.68610 -1.65153 2.13359

probit_b 289 0.18327 0.92935 52.96450 -2.54401 2.77741

Pearson Correlation Coefficients, N = 289

Prob > |r| under H0: Rho=0

probit_a probit_b

probit_a 1.00000 0.81173

<.0001

probit_b 0.81173 1.00000

<.0001

/udd/nhron/macrodevelopment/WilcxRank/testWilgoodnessfit.sas

]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]

Both models 10:12 Monday, March 3, 2008 11

------------------------------------------ base_agegrp=1 -------------------------------------------

The CORR Procedure

2 Variables: probit_a probit_b

Simple Statistics

Variable N Mean Std Dev Sum Minimum Maximum

probit_a 205 -0.35401 0.84019 -72.57224 -1.58920 1.93530

probit_b 205 -0.43153 0.88212 -88.46325 -2.45073 2.45073

Pearson Correlation Coefficients, N = 205

Prob > |r| under H0: Rho=0

probit_a probit_b

probit_a 1.00000 0.82457

<.0001

probit_b 0.82457 1.00000

<.0001

/udd/nhron/macrodevelopment/WilcxRank/testWilgoodnessfit.sas

]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]

Both models 10:12 Monday, March 3, 2008 12

------------------------------------------ base_agegrp=2 -------------------------------------------

The CORR Procedure

2 Variables: probit_a probit_b

Simple Statistics

Variable N Mean Std Dev Sum Minimum Maximum

probit_a 75 -0.51371 0.82337 -38.52851 -1.65153 1.15871

probit_b 75 -0.70387 0.87621 -52.79046 -2.54401 1.38151

Pearson Correlation Coefficients, N = 75

Prob > |r| under H0: Rho=0

probit_a probit_b

probit_a 1.00000 0.80816

<.0001

probit_b 0.80816 1.00000

<.0001

/udd/nhron/macrodevelopment/WilcxRank/testWilgoodnessfit.sas

]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]

Both models 10:12 Monday, March 3, 2008 13

Defference

base_ # of # of Theta +/- se Theta +/- se of Theta from

Obs agegrp Cases Controls for model1 for model2 2 models

1 1 285 205 0.689+/-0.024 0.725+/-0.023 -0.036+/-0.016

2 2 289 75 0.695+/-0.034 0.759+/-0.031 -0.063+/-0.021

/udd/nhron/macrodevelopment/WilcxRank/testWilgoodnessfit.sas

]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]]

Both models 10:12 Monday, March 3, 2008 14

# of # of Theta +/- se Theta +/- se Defference of

Cases Controls for model1 for model2 Theta from 2

Obs Summary Summary Summary Summary models Summary pvalue

1 574 280 0.691+/- 0.020 0.737+/- 0.018 -0.046+/- 0.013 .000304654

/udd/nhron/macrodevelopment/WilcxRank/testWilgoodnessfit.sas