Not every protein necessarily starts with methionine, however. Often this first amino acid will be removed in later processing of the protein. A tRNA charged with methionine binds to the translation start signal. The large subunit binds to the mRNA and the small subunit, and so begins elongation, the formation of the polypeptide chain. After the first charged tRNA appears in the A site, the ribosome shifts so that the tRNA is now in the P site. New charged tRNAs, corresponding the codons of the mRNA, enter the A site, and a bond is formed between the two amino acids. The first tRNA is now released, and the ribosome shifts again so that a tRNA carrying two amino acids is now in the P site. A new charged tRNA then binds to the A site. This process of elongation continues until the ribosome reaches what is called a stop codon, a triplet of nucleotides that signals the termination of translation. When the ribosome reaches a stop codon, no aminoacyl tRNA binds to the empty A site. This is the ribosome signal to break apart into its large and small subunits, releasing the new protein and the mRNA. Yet, this isn't always the end of the story. A protein will often undergo further modification, called post-translational modification. For example, it might be cleaved by a protein-cutting enzyme, called a protease, at a specific place or have a few of its amino acids altered.
Not actually. In archaea and some other organisms, even one fungal species can use other codon rather than AUG. But the amino acid necessarily coded is Met. exception to universal codon concept. this is violating the rule of AUG coding Met.
also in mitochondria the start codon differs.
for better understanding follow Lehninger biochemistry book (protein metabolism).
Almost all yes before post-translational modifications.
and less often after post-translational modifications.
First as they say in this publication (Hecht A., Glasgow J., Jaschke P.R., Bawazer L., Munson M.S., Cochran J., Endy D., Salit M. Measurements of translation initiation from all 64 codons in E. coli. Nucleic Acids Res. 2017; 45:3615–3626.) the translation can occur from pretty much any initial codon.
Then non-canonical start codons usually code for amino acids other than methionine, but when they act as START codons they code for Met (in Eukaryote) and fMet (in Prokaryote). In the publication mentioned above, they found one particular case for the codon ACG that seems to succeed to initiate the translation with AA Threonine. It might be explained by the fact that the Threonine is a Methionine cognate amino acid.
So, except few exception the first AA is always a Methionine... until post-translational modifications occur. Indeed post-translational modifications can excise the first AA, thus the Methionine can be removed.
Among others, that is discussed in those forums: https://biology.stackexchange.com/questions/56939/do-all-proteins-start-with-methionine