I have a multiple datasets but they are related to each other.

sample  condition
sample_1    WT_1
sample_2    WT_2
sample_3    WT_3
sample_4    Protein_1_KO_1
sample_5    Protein_1_KO_2
sample_6    Protein_1_KO_3
sample_7    WT_1
sample_8    WT_2
sample_9    WT_3
sample_10   Protein_2_KO_1
sample_11   Protein_2_KO_2
sample_12   Protein_2_KO_3
sample_13   WT_1
sample_14   WT_2
sample_15   WT_3
sample_16   Protein_1_2_KO_1
sample_17   Protein_1_2_KO_2
sample_18   Protein_1_2_KO_3
sample_19   Protein_3_KO_1
sample_20   Protein_3_KO_2
sample_21   Protein_3_KO_3

I have 21 paired end libraries. I have three different proteins that are deleted from the genome. My source of variation is Protein KO. can i design formula <- ~ condition or do i need something different. Can you please help me.

You have to remove the _1/_2/_3's from the condition column. The idea is that samples (rows) of the same group have the same identifier there. The vignette covers that, please read it.