Dear Sofia,

please excuse the misunderstandings caused by my terminology.
You can rename all lowercase "fc" below
to normalized_signal or
log transformed ratio of signal to input.
Only the 2 new "FC" are log2foldchanges.


At
http://genomics-lab.fleming.gr/fleming/DKlab/run392/run402/exonGencode.groups.rpkm-v2.csv
you can load the RPKM analysis per exon for the liver samples with corrected q-values (FDR).
With a FDR <= 0.1 (which is common practice), there are
804 exons for M2RIP_WTvsInputWT_vs_M2RIP_Tg6105vsInputTg6105 and
73 exons for HuR_RIPC57r_vs_IgG_RIPC57 with a significant signal.
I've also added the foldchanges in the last 2 columns.

The definition of the exon in col. 1 to 6:
 [1] "Geneid"                                             
 [2] "Chr"                                                
 [3] "Start"                                              
 [4] "End"                                                
 [5] "Strand"                                             
 [6] "Length"

The non-input RPKM values in col 7 to 18: 
 [7] "DKR13r_M2RiPWt_1M"                                  
 [8] "DKR14r_M2RiPWt_2M"                                  
 [9] "DKR15r_M2RiPWt_3F"                                  
[10] "DKR16_M2RiPTg6105_1M"                               
[11] "DKR17r_M2RiPTg6105_2M"                              
[12] "DKR18_M2RiPTg6105_3F"                               
[13] "DKR19_HuRRiPC57_1M"                                 
[14] "DKR21_HuRRiPC57_3F"                                 
[15] "DKR20_HURRIPC57_2M"                                 
[16] "DKR22_IgGRiPC57_1M"                                 
[17] "DKR23r_IgGRiPC57_2M"                                
[18] "DKR24r_IgGRiPC57_3F"                                

The input RPKM values (averaged over the 3 replicates) in col 19 to 21
[19] "iWT"                                                
[20] "iTg6105"                                            
[21] "iC57"                                               

The log2 foldchanges as defined below in (a) in col 22 to 30
[22] "M2RIP_WT1_fc"                                       
[23] "M2RIP_WT2_fc"                                       
[24] "M2RIP_WT3_fc"                                       
[25] "M2RIP_Tg6105_1_fc"                                  
[26] "M2RIP_Tg6105_2_fc"                                  
[27] "M2RIP_Tg6105_3_fc"                                  
[28] "HuR_RIPC57_1_fc"                                    
[29] "HuR_RIPC57_2_fc"                                    
[30] "HuR_RIPC57_3_fc"                                    

p-values and correced p-vaules (FDR or q-values) of a  t.test for M2RIP_WTvsInputWT vs M2RIP_Tg6105vsInputTg610
[31] "M2RIP_WTvsInputWT_vs_M2RIP_Tg6105vsInputTg6105_pval"
[32] "M2RIP_WTvsInputWT_vs_M2RIP_Tg6105vsInputTg6105_qval"

The log2 foldchanges as defined below in (b)

[33] "IgG_RIPC57_1_fc"                                    
[34] "IgG_RIPC57_2_fc"                                    
[35] "IgG_RIPC57_3_fc"                                    

p-values and correced p-vaules (FDR or q-values) of a  t.test for HuR_RIPC57rVSinput_vs_IgG_RIPC57VSinput
[36] "HuR_RIPC57r_vs_IgG_RIPC57_pval"                     
[37] "HuR_RIPC57r_vs_IgG_RIPC57_qval"                     

log2 fold changes of normalized signals, calculated using (c):
[38] "M2RIP_WTvsInputWT_vs_M2RIP_Tg6105vsInputTg6105_FC"  
[39] "HuR_RIPC57r_vs_IgG_RIPC57_FC"                       

(a)
# M2RIP_WT vs InputWT
r$M2RIP_WT1_fc=log2((1+r$DKR13r_M2RiPWt_1M)/(1+r$iWT));
r$M2RIP_WT2_fc=log2((1+r$DKR14r_M2RiPWt_2M)/(1+r$iWT));
r$M2RIP_WT3_fc=log2((1+r$DKR15r_M2RiPWt_3F)/(1+r$iWT));

# M2RIP_Tg6105 vs InputTg6105
r$M2RIP_Tg6105_1_fc=log2((1+r$DKR16_M2RiPTg6105_1M )/(1+r$iTg6105));
r$M2RIP_Tg6105_2_fc=log2((1+r$DKR17r_M2RiPTg6105_2M)/(1+r$iTg6105));
r$M2RIP_Tg6105_3_fc=log2((1+r$DKR18_M2RiPTg6105_3F )/(1+r$iTg6105));

# HuR_RIPC57 vs InputC57
r$HuR_RIPC57_1_fc=log2((1+r$DKR19_HuRRiPC57_1M)/(1+r$iC57));
r$HuR_RIPC57_2_fc=log2((1+r$DKR21_HuRRiPC57_3F)/(1+r$iC57));
r$HuR_RIPC57_3_fc=log2((1+r$DKR20_HURRIPC57_2M)/(1+r$iC57));

(b)
r$IgG_RIPC57_1_fc=log2((1+r$DKR22_IgGRiPC57_1M)/(1+r$iC57));
r$IgG_RIPC57_2_fc=log2((1+r$DKR23r_IgGRiPC57_2M)/(1+r$iC57));
r$IgG_RIPC57_3_fc=log2((1+r$DKR24r_IgGRiPC57_3F)/(1+r$iC57));

(c)
a=apply(cbind(r$M2RIP_WT1_fc,r$M2RIP_WT2_fc,r$M2RIP_WT3_fc),1,mean)
b=apply(cbind(r$M2RIP_Tg6105_1_fc,r$M2RIP_Tg6105_2_fc,r$M2RIP_Tg6105_3_fc),1,mean)
r$M2RIP_WTvsInputWT_vs_M2RIP_Tg6105vsInputTg6105_FC=a-b

a=apply(cbind(r$HuR_RIPC57_1_fc,r$HuR_RIPC57_2_fc,r$HuR_RIPC57_3_fc),1,mean)
b=apply(cbind(r$IgG_RIPC57_1_fc,r$IgG_RIPC57_2_fc,r$IgG_RIPC57_3_fc),1,mean)
r$HuR_RIPC57r_vs_IgG_RIPC57_FC=a-b


BW,
Martin