I have come across an online ppt slide of the bioinformatic algorithm where it is said that first and the last amino acid Ramachandran angle is not necessary to tell all its internal coordinates. Is it really not necessary for any simulations while predicting the 3D structure of proteins? For instance, if we have a protein chain of N amino acid sequence, there will be 2N dihedral angles. Of which we need to specify only the 2N-2 (apart from the first and the last amino acid dihedral angles) dihedral angles. Why the two angles are ignored? Let's assume all the bond angles and bond lengths are provided for the entire chain.
I have seen this statement in the following ppt, slide number -9, titled representation of internal coordinates, under bullet point 3.