Dear all,
This is a real dataset. So I have scrambled a bit the biomarker names. Bear with me.
The objective is to use -meta- to plot regression coefficients and the corresponding p-value.
I am aware there are other alternatives such as -coefplot-, scatter/rcap, -metan- etc. But the -meta- graphs seem a bit nicer.
The problem: -meta forestplot- fails to align well strings. Any suggestions on how to solve the problem?
See a real example below.
All the best,
Tiago
This is a real dataset. So I have scrambled a bit the biomarker names. Bear with me.
The objective is to use -meta- to plot regression coefficients and the corresponding p-value.
I am aware there are other alternatives such as -coefplot-, scatter/rcap, -metan- etc. But the -meta- graphs seem a bit nicer.
The problem: -meta forestplot- fails to align well strings. Any suggestions on how to solve the problem?
See a real example below.
All the best,
Tiago
Code:
*! prepare the dataset
clear
input str35 biomarker N mean lower upper
"Sadasd (g/L)" 67 -2.005361 -6.76237 2.751647
"iclmZa citponosá ({&mu}g/dL)" 103 0.0344073 -0.0984404 0.1672551
"oáZrinocn riiu ({&mu}g/L)" 83 0.0073011 -0.002171 0.0167733
"meC tiolbáaospcr ({&mu}g/dL)" 103 0.0683467 -0.0063204 0.1430137
"tugae sn olerCabtono ({&mu}g/L)" 104 0.0061882 -0.0166123 0.0289888
"rt eooab nClteouagsn ({&mu}g/L)" 83 -0.0164623 -0.0703007 0.0373761
":ozoaRzãniceocbr" 103 5.932471 -1.133889 12.99883
"omipesêáotaScni ll ({&mu}g/L)" 104 0.0046533 -0.058296 0.0676026
" ltS aoeno aientsgulnêo ({&mu}g/L)" 104 0.0158911 -0.0262121 0.0579943
"e inutnSoelo sgltêaao n ({&mu}g/L)" 83 -0.0074458 -0.1571615 0.14227
"innua eonArgssoê ({&mu}g/L)" 104 1.351535 -2.159285 4.862354
"náosunoriAriê ir ({&mu}g/L)" 83 0.039656 -0.0359255 0.1152374
"oendáaguoCmns I ({&mu}g/L)" 104 0.9043854 -1.608069 3.41684
"Cmidoná iráoiur ({&mu}g/L)" 83 -9.887843 -33.37931 13.60362
"iCirmnourdáioá ({&mu}g/L)" 104 0.0330633 -0.0170313 0.083158
"ui rirMrráoúioecn ({&mu}g/L)" 83 -0.3591603 -1.359202 0.6408809
end
gene p_value = round(runiform(),0.01)
format %9.2f p_value
label variable biomarker `"Biomarker"'
label variable p_value `"{it:P}"'
*! set data for meta-analysis
meta set mean lower upper, studylabel(biomarker)
*! plot the results using the standard approach
*! results look good - well aligned at position -99 (see the red line)
meta forest, cibind(parentheses) overallopts(size(zero) color(white)) ///
nowmarkers noomarker markeropts(msize(6-pt) msymbol(circle)) noohetstats ///
noghetstats noosigtest noohomtest nogwhomtests nogbhomtests nonotes ///
xline(0, lpatter(solid) lcolor(black*0.5)) columnopts(_esci, supertitle({&beta}) ///
title(95% CI)) hruleopts(lcolor(none)) coltitleopts(justification(center)) xline(-99)
*! plot the results + other information
*! results are no longer aligned - see the red line at position -94
meta forest biomarker _plot _esci p_value, cibind(parentheses) ///
overallopts(size(zero) color(white)) nowmarkers noomarker ///
markeropts(msize(6-pt) msymbol(circle)) noohetstats noghetstats ///
noosigtest noohomtest nogwhomtests nogbhomtests ///
nonotes xline(0, lpatter(solid) lcolor(black*0.5)) ///
columnopts(_esci, supertitle({&beta}) title(95% CI)) ///
hruleopts(lcolor(none)) coltitleopts(justification(center)) xline(-94)
Comment