As there has to be some kind of a minimal overlap betwen the oligo-d(T) and the poly(A)
tail, this is incorrect. This will most likely introduce a slight shift in tail length estimations for real data.
46062bb introduces a min_offset parameter to simulate_reads that already allows to change this behaviour for the simulations, but the model couldn't handle this as of now.
Right now we allow for binding of the oligo-d(T) anywhere in the poly(A) tail. This is implicitely used in
prob_d_given_L(twice), as well as inprob_d_given_L_weighted(also twice).As there has to be some kind of a minimal overlap betwen the oligo-d(T) and the poly(A) tail, this is incorrect. This will most likely introduce a slight shift in tail length estimations for real data.
46062bb introduces a
min_offsetparameter tosimulate_readsthat already allows to change this behaviour for the simulations, but the model couldn't handle this as of now.