rubabel

require 'rubabel'

# by default, reads in smiles strings
serine = Rubabel["C(C(C(=O)O)N)O"]
# more formally:
serine = Rubabel::Molecule.from_string("C(C(C(=O)O)N)O")

# also any other format openbabel supports, for example inchi
serine = Rubabel["InChI=1S/C3H7NO3/c4-2(1-5)3(6)7/h2,5H,1,4H2,(H,6,7)", :inchi]

# from an InChI Key or other id (such as LipidMaps ID) [requires internet]
serine = Rubabel["MTCFGRXMJLQNBG-REOHCLBHSA-N", :inchikey]

Find out all the formats Rubabel supports (hash is format key pointing to the description):

hash = Rubabel.in_formats
hash = Rubabel.out_formats
hash = Rubabel.id_formats # molecule retrieved by web-lookup

Rubabel.foreach("file.sdf") do |mol|
  puts mol.exact_mass
end

# return an array of every unique atom type in the file
uniq_atom_types = Rubabel.foreach("file.mol").flat_map {|mol| mol.map(&:type) }.uniq

mol = Rubabel::Molecule.from_file("file.sdf")
# handles gzipped files seamlessly:
mol = Rubabel::Molecule.from_file("file.sdf.gz") 
mol = Rubabel.molecule_from_file("file.sdf") # alternative

# explicit format for uninformative/wrong extensions:
mol = Rubabel::Molecule.from_file("file", :sdf)

mol = Rubabel["OCC"] # ethanol

mol.to_s       # canonical smiles -> "CCO"
mol.csmiles    # same thing

mol.to_s(:smi) # smiles -> "OCC"
mol.smiles     # same thing

can_smiles_string = mol.write_string # -> "CCO\t\n"
mol.title = "ethanol"
can_smiles_string = mol.write(:can)  # -> "CCO\tethanol\n"

pdb_string = mol.write(:pdb)

# write to a smiles file
mol.write("file.smi")
mol.write_file("file.smi")

File.open("somefile.pdb", 'w') do |out|
  molecules.each {|mol| out.print mol.write(:pdb) }
end

mol = Rubabel["NCC(O)C(=O)O"]
mol.write("file.svg")

# must have imagemagick ('convert' command) and mini_magick gem installed
mol.write("file.png") 

mol = Rubabel["NCC(O)C(=O)O"]
mol.each_match("CO") do |match|
  # match is just an array of atoms that matched
  match.first.el # => :C
  match.last.el # => :O
end

# matches returns all the matches in an array
all_matches = mol.matches("CO")
# all the match routines take a boolean to alter uniqueness
all_matches = mol.matches("CO", false) # some matches may not be uniq

bonds = mol.matches("CO").map {|c, o| c.get_bond(o) }
mol.split(*bonds)  # splits between every carbon single bonded to oxygen

mol = Rubabel["OCC"]
# adds a carbon, then an oxygen to the previous carbon
mol << 6 << 8    # #<Mol "OCCCO">
mol << :C << :O  # same thing

# add an ethyl group specifically to second atom (the first carbon)
mol = Rubabel["OCC"]
mol[1] << :C << :C

# add a vinyl group to second carbon (use method notation and parenthesis 
# because we are going to specify 2 arguments (the bond order):
( mol[1] << :C).<<(:C, 2)

# delete an atom:
mol = Rubabel["NCO"]
mol.delete(mol[0])  
# mol.to_s -> #<Mol CO>

# delete a bond:
bond = mol[0].get_bond(mol[1])
mol.delete(bond)  
# mol.to_s ->  #<Mol C.O>

mol = Rubabel["CC"]
bond = mol[0].get_bond(mol[1])  # get the bond you want
bond = mol[0].bonds.first       # alternate way to get at bonds

bond += 2    # now it is a triple bond
bond.dec!    # now a double bond
bond -= 1    # now a single bond
bond.inc!(2) # back to a triple bond