issues
search
rapodaca
/
dialect
Documenting a subset of the SMILES language.
MIT License
13
stars
0
forks
source link
Discussion Section
#14
Closed
rapodaca
closed
2 years ago
rapodaca
commented
3 years ago
Scope, limitations, observations. Maybe some more context.
benefits
compact
examples with byte counts
compare with molfile
use in REPLs
Jupyter
hand-codable
easy to learn
designed for lossless (de)serialization
c.f. InChI
handles most of organic chemistry
tradeoffs
non-representable entities
non-VB examples
organometallics, homotropylium cation, dative bonding
non-tetrahedral stereochemistry
e.g., TB, OH, lone-pair tetrahedral
conformational restriction other than (E)/(Z) double bonds
PPBs are error-prone
compatibility
SMILES itself is not well-defined
gather public documentation on syntax/semantics (just the docs)
Daylight, OpenSMILES, CDK, OpenBabel, RDKit, Jchem, OE, etc.
in case of conflict among tools, choose ease of implementation
unsupported
some selections (e.g. [se])
the "aromatic" bond (
:
)
extreme charges (< -9, > 9)
arbitrary element symbols
extensions
extension field
can be used to encode application-specific information as integer
limited range (0-9999)
can be used together with metadata
versioning
breaks compatibility, but maybe metadata format
metaformats - maybe
in-line vs out-of-line
leverage extension
expanding range of selectable atoms
additional configuration classes (OH, TB, etc)
canonicalization
"preferred format"
atomic numbering
outlook
detailed spec opens new paths
reference implementation
to be reported
validation suites
improve data quality by detecting syntax/semantics differences
writing better implementations
reading/writing sections
performance benchmarks
apples/apples comparisons using the same protocols
faster processing
standardization efforts
more detailed, structured source material to draw from
select, rather than develop elements of standard
better line notations
clearly-delineated scope and limitations
lots of room for improvement
may or may not happen through dialect extensions
Scope, limitations, observations. Maybe some more context.
:
)