**Instructions:**  
Analyze the given molecule represented by a SMILES string to determine the total number of alpha-amino acid substructures present in the molecule.

- Assign unique indices to all non-hydrogen atoms in the molecule based on their order of appearance in the SMILES string (left to right), starting from 0.
- Identify alpha-amino acid substructures, defined as a contiguous fragment of the form N-C_alpha(-R)-C(=O)O, where:
  - N is a amino nitrogen (NH2),
  - C_alpha is the carbon atom directly bonded to this nitrogen and to the carboxyl carbon,
  - C_alpha is the carbon atom directly bonded to this nitrogen and to the carboxyl carbon, and also connected by a single bond to a side chain R (it might be a single hydrogen),
  - C(=O)O is a carboxyl or carboxylate group (COOH or COO-) directly attached to C_alpha.
- Count the total number of alpha-amino acid substructures.
- For each alpha-amino acid substructure, provide the indices of the five backbone atoms in the following order:
  1. Amino nitrogen (N)
  2. Alpha carbon (C_alpha)
  3. Carboxyl carbon (C)
  4. Carbonyl oxygen (O connected by a double bond to C)
  5. Hydroxyl / single-bonded oxygen (O connected by a single bond to C)
- Present the results in the following structured format:
  1. Total number of alpha-amino acid substructures: Enclose this integer within <count> and </count>. 
  2. Indices of alpha-amino acid backbone atoms: Enclose the list of atom indices for each alpha-amino acid substructure within <atom_indices> and </atom_indices>. Each substructure should be represented as a separate list of five indices. 
- For example, if two alpha-amino acid substructures are found, the output should be formatted as follows:
  1. <count>2</count>
  2. <atom_indices> [0, 1, 2, 3, 4], [5, 6, 7, 8, 9] </atom_indices>

**SMILES string:** `{smiles}`