dear all,
I have a dataset composed by different classes: I have two main classes,
"active (Act.)" and "latent (Lat.)", furher subdivided in
the gene
subclasses "IP10" and "MIG" and then into 7 stimulus
sub-subclasses "ESAT6"
to "PHA".
Would it be possible to test in a direct way the statistical differences
between the response variable ("ratio") of the classes active and
latent
values for each of the sub-subclasses and stratified by the subclass gene?
that is: how can i find differences for, let's say, the ESAT6 ratios
between active and latent for the gene IP10, then for the ESAT6 active
versus latent for the MIG gene; then reiterate for the CFP10 for IP10,
CFP10 for MIG and so on and so forth.
the data looks like:
TB gene stimulus ratio
Act. IP10 ESAT6 0.752020287
Act. IP10 CFP10 0.893628674
Act. IP10 Rv3615c 0.958108603
Act. IP10 Rv2654 0.319492309
Act. IP10 Rv3879 0.074381059
Act. IP10 Rv3873 0.874629325
Act. IP10 PHA 0.308093623
Act. MIG ESAT6 0.398817499
Act. MIG CFP10 0.022455057
Act. MIG Rv3615c 0.186013561
Act. MIG Rv2654 0.040591894
Act. MIG Rv3879 0.716143409
Act. MIG Rv3873 0.271241311
Act. MIG PHA 0.154207839
Lat. IP10 ESAT6 0.579278956
Lat. IP10 CFP10 0.553393777
Lat. IP10 Rv3615c 0.918602425
Lat. IP10 Rv2654 0.56821826
Lat. IP10 Rv3879 0.429991707
Lat. IP10 Rv3873 0.938468864
Lat. IP10 PHA 0.621095162
Lat. MIG ESAT6 0.486344894
Lat. MIG CFP10 0.288455361
Lat. MIG Rv3615c 0.602348528
Lat. MIG Rv2654 0.717395188
Lat. MIG Rv3879 0.178745118
Lat. MIG Rv3873 0.330885875
Lat. MIG PHA 0.327117342
Many thanks for any possible help.
Best regards,
Luigi
[[alternative HTML version deleted]]