The PSMOOTH Procedure

Example

Suppose you want to test the 16 markers represented in the following data for association with a disease by using the genotype case-control and trend tests in PROC CASECONTROL. You are concerned about the multiple hypothesis testing issue, and so you also want to run PROC PSMOOTH on the output data set from PROC CASECONTROL in order to eliminate the number of false positives found using the individual $\text{[math]}$ -values from the marker-trait association tests.

data in;
   input affected (m1-m16) ($);
   datalines;
1 1/2 2/2 2/2 2/2 1/1 2/2 1/2 1/2 1/1 1/2 1/2 2/2 2/2 2/2 2/2 1/2
1 1/2 1/1 1/2 1/2 1/1 1/1 1/2 1/1 1/2 1/2 1/1 2/2 1/1 1/2 1/1 1/2
1 1/1 2/2 1/2 1/2 1/1 1/2 1/1 1/2 1/2 2/2 2/2 1/2 1/2 1/2 2/2 1/2
1 1/1 1/2 2/2 1/2 1/2 1/1 1/2 1/2 1/2 1/1 1/1 1/2 2/2 1/2 1/1 1/1
1 1/2 1/1 1/1 1/2 2/2 1/1 1/1 1/2 1/1 2/2 1/2 2/2 2/2 2/2 1/2 1/1

   ... more lines ...   

0 2/2 1/1 1/2 1/1 1/2 1/2 1/2 2/2 1/1 1/2 1/1 1/1 1/1 2/2 1/1 1/2
;

Note that the columns marker1-marker16 contain genotypes at each of the markers, so the GENOCOL option must be used in PROC CASECONTROL as follows to correctly read in the data.

proc casecontrol data=in outstat=cc_tests genotype trend genocol;
   trait affected;
   var m1-m16;
run;

proc psmooth data=cc_tests simes fisher tpm bw=2 adjust=sidak 
             out=adj_p;
   var ProbGenotype ProbTrend;
   id Locus;
run;

  
proc print data=adj_p heading=h; 
run;

This code modifies the $\text{[math]}$ -values contained in the output data set from PROC CASECONTROL, first by smoothing the $\text{[math]}$ -values by using Simes’ method, Fisher’s method, and the TPM with a bandwidth of 2, then by applying Sidak’s multiple testing adjustment to the smoothed $\text{[math]}$ -values.

Figure 11.1 PROC PSMOOTH Output Data Set

Obs	Locus	ProbGenotype	ProbGenotype_S2	ProbGenotype_F2	ProbGenotype_T2	ProbTrend	ProbTrend_S2	ProbTrend_F2	ProbTrend_T2
1	m1	0.61481	0.84871	0.96719	0.83858	0.32699	1.00000	1.00000	0.91474
2	m2	0.03711	0.92355	0.97753	0.91260	0.84733	1.00000	1.00000	0.96248
3	m3	0.57096	0.96252	0.98449	0.95280	0.57628	1.00000	0.99986	0.98348
4	m4	0.34059	0.96252	0.80318	0.95280	0.23932	1.00000	1.00000	0.98348
5	m5	0.35600	0.99999	0.99858	0.98348	0.16135	0.99998	0.99979	0.98348
6	m6	0.12375	0.99999	0.99861	0.98348	0.85742	0.99981	0.99807	0.98348
7	m7	0.41529	1.00000	0.99962	0.98348	0.29694	0.99994	0.99961	0.98348
8	m8	0.57360	1.00000	0.99997	0.98348	0.33141	0.99994	0.99999	0.98348
9	m9	0.47332	1.00000	1.00000	0.98348	0.36231	0.99925	0.99902	0.98348
10	m10	0.59452	0.05946	0.41423	0.25944	0.31242	0.01520	0.06303	0.11769
11	m11	0.44085	0.05946	0.02931	0.01550	0.35299	0.01520	0.01179	0.01454
12	m12	0.00076	0.05946	0.00036	0.00017	0.00019	0.01345	0.00005	0.00005
13	m13	0.00911	0.05946	0.00052	0.00017	0.03301	0.01345	0.00011	0.00005
14	m14	0.00160	0.05946	0.00008	0.00002	0.00034	0.01345	0.00001	0.00000
15	m15	0.94287	0.09744	0.00570	0.00138	0.86176	0.02144	0.00153	0.00044
16	m16	0.04264	0.07395	0.05720	0.01902	0.01207	0.01612	0.00519	0.00223

Figure 11.1 displays the original and modified $\text{[math]}$ -values.