Add vectors described in Sneath (1966)

docs/index.rst

@@ -79,8 +79,9 @@ A non-exhaustive list of available features:
   - `FASGAI vectors <https://doi.org/10.1111/j.1747-0285.2008.00641.x>`_
   - `Kidera factors <https://doi.org/10.1007/BF01025492>`_
   - `MS-WHIM scores <https://doi.org/10.1021/ci980211b>`_
-  - `PCP descriptors <https://doi.org/10.1007/s00894-001-0058-5>`_.
+  - `PCP descriptors <https://doi.org/10.1007/s00894-001-0058-5>`_
   - `ProtFP descriptors <https://doi.org/10.1186/1758-2946-5-41>`_
+  - `Sneath vectors <10.1016/0022-5193(66)90112-3>`_
   - `ST-scales <https://doi.org/10.1007/s00726-009-0287-y>`_
   - `T-scales <https://doi.org/10.1016/j.molstruc.2006.07.004>`_
   - `VHSE-scales <https://doi.org/10.1002/bip.20296>`_

peptides/__init__.py

@@ -93,6 +93,13 @@ class ProtFPDescriptors(typing.NamedTuple):
     protfp8: float
 
 
+class SneathVectors(typing.NamedTuple):
+    sv1: float
+    sv2: float
+    sv3: float
+    sv4: float
+
+
 class STScales(typing.NamedTuple):
     st1: float
     st2: float
@@ -1601,6 +1608,48 @@ class Peptide(typing.Sequence[str]):
             )
         return ProtFPDescriptors(*out)
 
+    def sneath_vectors(self) -> SneathVectors:
+        """Compute the Sneath vectors for a protein sequence.
+
+        These vectors were obtained in Sneath (1996) by running PCA on the
+        `ϕ coefficient <https://en.wikipedia.org/wiki/Phi_coefficient>`_
+        to explain the dissimilarity between the 20 natural amino acids
+        based on binary state encoding of 134 physical and chemical
+        properties (such as presence/absence of a —CH₃ group, step-wise
+        optical rotation, etc.).
+
+        Returns:
+            `peptides.SneathVectors`: The computed average of Sneath vectors
+            of all the amino acids in the peptide. *SV1* appears to
+            represent mainly aliphatic properties, *SV2* may model the
+            number of reactive groups, *SV3* the aromatic properties, but
+            *SV4* has no certain interpretation.
+
+        Example:
+            >>> peptide = Peptide("QWGRRCCGWGPGRRYCVRWC")
+            >>> for i, fp in enumerate(peptide.sneath_vectors()):
+            ...     print(f"SV{i+1:<3} {fp: .4f}")
+            SV1    0.1962
+            SV2    0.0466
+            SV3    0.0405
+            SV4    0.0277
+
+        References:
+            - Sneath, P. H. A.
+              *Relations between Chemical Structure and Biological Activity
+              in Peptides*.
+              Journal of Theoretical Biology. Nov 1996;12(2):157–95.
+              doi:10.1016/0022-5193(66)90112-3. PMID:4291386.
+
+        """
+        out = array.array("d")
+        for i in range(len(tables.SNEATH)):
+            scale = tables.SNEATH[f"SV{i+1}"]
+            out.append(
+                sum(scale.get(aa, 0) for aa in self.sequence) / len(self.sequence)
+            )
+        return SneathVectors(*out)
+
     def st_scales(self) -> STScales:
         """Compute the ST-scales of a protein sequence.
 
@@ -1775,7 +1824,9 @@ class Peptide(typing.Sequence[str]):
         "F": fasgai_vectors,
         "KF": kidera_factors,
         "MSWHIM": ms_whim_scores,
+        "E": pcp_descriptors,
         "ProtFP": protfp_descriptors,
+        "SV": sneath_vectors,
         "ST": st_scales,
         "T": t_scales,
         "VHSE": vhse_scales,

peptides/data/sneath/SV1.csv

+A,0.239
+R,0.211
+N,0.249
+D,0.171
+C,0.220
+Q,0.187
+E,0.260
+G,0.160
+H,0.205
+I,0.273
+L,0.281
+K,0.228
+M,0.253
+F,0.234
+P,0.165
+S,0.236
+T,0.213
+W,0.183
+Y,0.193
+V,0.255

peptides/data/sneath/SV2.csv

+A,0.330
+R,-0.176
+N,-0.233
+D,-0.371
+C,0.074
+Q,-0.409
+E,-0.254
+G,0.370
+H,-0.078
+I,0.149
+L,0.129
+K,-0.075
+M,-0.092
+F,-0.011
+P,0.370
+S,0.022
+T,0.136
+W,-0.011
+Y,-0.138
+V,0.245

peptides/data/sneath/SV3.csv

+A,-0.110
+R,0.079
+N,-0.136
+D,-0.285
+C,-0.184
+Q,-0.246
+E,-0.067
+G,-0.073
+H,0.320
+I,0.001
+L,-0.008
+K,0.049
+M,-0.041
+F,0.438
+P,-0.016
+S,-0.153
+T,-0.208
+W,0.493
+Y,0.381
+V,-0.155

peptides/data/sneath/SV4.csv

+A,-0.062
+R,-0.167
+N,0.166
+D,-0.079
+C,0.380
+Q,-0.184
+E,-0.025
+G,-0.017
+H,0.056
+I,-0.309
+L,-0.264
+K,-0.371
+M,0.077
+F,0.074
+P,-0.036
+S,0.470
+T,0.348
+W,0.050
+Y,0.220
+V,-0.212