This project is to extract electrostatic features from a protein structure. Proposed stages are:
- Stage 1. Extract proposed electrostatic features an MCCE working directory.
- Stage 2. Extract electrostatic features from a pdb file directly from MCCE-ML module.
- Stage 3. Integrate electrostatic feature to PClass classifier
Total features: 83 standard features. Additional custom numeric features can be
appended from an optional mcce-features.txt file in each MCCE folder.
The feature set summarizes protein electrostatics from MCCE output files. Charges
are taken at pH 6.0, 7.0, and 8.0 from sum_crg.out, solvent accessibility from acc.res,
pKa values from pK.out, and residue coordinates from step1_out.pdb.
Surface residues are residues with solvent-accessible surface area fraction
greater than 0.1. Acidic residues are ASP, GLU, and CTR; basic residues are LYS,
ARG, HIS, NTR, and NTG.
Use mcce-features.txt when you already have extra numeric descriptors for a
protein and want mcce-features to carry them into the same output table as the
standard electrostatic features. This is useful for adding values from another
program, manual curation, or a separate analysis step without changing the
mcce-features source code.
Place the file in the MCCE working directory being extracted. The filename must be exactly:
mcce-features.txt
Each non-empty line defines one additional feature as:
feature_name: value
The feature name becomes the output column name, and the value must be numeric. Multiple feature/value pairs are allowed.
Example:
folding_energy: -12.7
sas_to_volume: 2.3
When the file is present, MCCEFeatureExtractor reads these pairs and appends
them after the 83 standard feature columns. In batch output, custom columns are
included when any processed folder defines them; folders that do not define a
given custom feature leave that value blank.
If a line is malformed, has an empty feature name, or has a non-numeric value, that line is skipped and a warning is logged. If a custom feature uses the same name as a standard feature, it is skipped to avoid overwriting the standard calculation.
net_charge: Total protein charge at pH 7.0 from theNet_Chargerow insum_crg.out.isoelectric_point: Estimated pH where total protein charge crosses zero.acid_fraction_all_residues: Fraction of all residues that are acidic.glu_fraction_all_residues: Fraction of all residues that are GLU.asp_fraction_all_residues: Fraction of all residues that are ASP.base_fraction_all_residues: Fraction of all residues that are basic.lys_fraction_all_residues: Fraction of all residues that are LYS.arg_fraction_all_residues: Fraction of all residues that are ARG.acid_to_base_ratio: Acidic-to-basic residue count ratio, using 999.0 when acids exist but bases do not.
acid_big_pka_shift_fraction_all_residues: Fraction of all residues that are acidic residues with large pKa shifts.glu_big_pka_shift_fraction_all_residues: Fraction of all residues that are GLU residues with large pKa shifts.asp_big_pka_shift_fraction_all_residues: Fraction of all residues that are ASP residues with large pKa shifts.base_big_pka_shift_fraction_all_residues: Fraction of all residues that are basic residues with large pKa shifts.lys_big_pka_shift_fraction_all_residues: Fraction of all residues that are LYS residues with large pKa shifts.arg_big_pka_shift_fraction_all_residues: Fraction of all residues that are ARG residues with large pKa shifts.acid_big_pka_shift_fraction_acids_only: Fraction of acidic residues with large pKa shifts.glu_big_pka_shift_fraction_acids_only: Fraction of acidic residues that are GLU with large pKa shifts.asp_big_pka_shift_fraction_acids_only: Fraction of acidic residues that are ASP with large pKa shifts.base_big_pka_shift_fraction_bases_only: Fraction of basic residues with large pKa shifts.lys_big_pka_shift_fraction_bases_only: Fraction of basic residues that are LYS with large pKa shifts.arg_big_pka_shift_fraction_bases_only: Fraction of basic residues that are ARG with large pKa shifts.mean_abs_pka_shift: Mean absolute pKa shift among acidic and basic residues.max_abs_pka_shift: Maximum absolute pKa shift among acidic and basic residues.
surface_net_charge_6.0: Net charge of solvent-exposed residues at pH 6.0.surface_acid_to_base_ratio_6.0: Exposed negative-to-positive charge ratio at pH 6.0.surface_positive_charge_density_6.0: Exposed positive charge per exposed SASA at pH 6.0.surface_negative_charge_density_6.0: Exposed negative charge magnitude per exposed SASA at pH 6.0.surface_net_charge_7.0: Net charge of solvent-exposed residues at pH 7.0.surface_acid_to_base_ratio_7.0: Exposed negative-to-positive charge ratio at pH 7.0.surface_positive_charge_density_7.0: Exposed positive charge per exposed SASA at pH 7.0.surface_negative_charge_density_7.0: Exposed negative charge magnitude per exposed SASA at pH 7.0.surface_net_charge_8.0: Net charge of solvent-exposed residues at pH 8.0.surface_acid_to_base_ratio_8.0: Exposed negative-to-positive charge ratio at pH 8.0.surface_positive_charge_density_8.0: Exposed positive charge per exposed SASA at pH 8.0.surface_negative_charge_density_8.0: Exposed negative charge magnitude per exposed SASA at pH 8.0.
buried_net_charge_6.0: Net charge of buried residues at pH 6.0.buried_acid_to_base_ratio_6.0: Buried negative-to-positive charge ratio at pH 6.0.buried_positive_charge_density_6.0: Buried positive charge per buried-residue SASA at pH 6.0.buried_negative_charge_density_6.0: Buried negative charge magnitude per buried-residue SASA at pH 6.0.buried_net_charge_7.0: Net charge of buried residues at pH 7.0.buried_acid_to_base_ratio_7.0: Buried negative-to-positive charge ratio at pH 7.0.buried_positive_charge_density_7.0: Buried positive charge per buried-residue SASA at pH 7.0.buried_negative_charge_density_7.0: Buried negative charge magnitude per buried-residue SASA at pH 7.0.buried_net_charge_8.0: Net charge of buried residues at pH 8.0.buried_acid_to_base_ratio_8.0: Buried negative-to-positive charge ratio at pH 8.0.buried_positive_charge_density_8.0: Buried positive charge per buried-residue SASA at pH 8.0.buried_negative_charge_density_8.0: Buried negative charge magnitude per buried-residue SASA at pH 8.0.
largest_positive_patch_area_6.0: SASA of the largest positive charge patch at pH 6.0.largest_negative_patch_area_6.0: SASA of the largest negative charge patch at pH 6.0.largest_positive_patch_charge_6.0: Total charge of the largest positive patch at pH 6.0.largest_negative_patch_charge_6.0: Total charge of the largest negative patch at pH 6.0.largest_positive_patch_density_6.0: Charge density of the largest positive patch at pH 6.0.largest_negative_patch_density_6.0: Charge density magnitude of the largest negative patch at pH 6.0.largest_positive_patch_area_7.0: SASA of the largest positive charge patch at pH 7.0.largest_negative_patch_area_7.0: SASA of the largest negative charge patch at pH 7.0.largest_positive_patch_charge_7.0: Total charge of the largest positive patch at pH 7.0.largest_negative_patch_charge_7.0: Total charge of the largest negative patch at pH 7.0.largest_positive_patch_density_7.0: Charge density of the largest positive patch at pH 7.0.largest_negative_patch_density_7.0: Charge density magnitude of the largest negative patch at pH 7.0.largest_positive_patch_area_8.0: SASA of the largest positive charge patch at pH 8.0.largest_negative_patch_area_8.0: SASA of the largest negative charge patch at pH 8.0.largest_positive_patch_charge_8.0: Total charge of the largest positive patch at pH 8.0.largest_negative_patch_charge_8.0: Total charge of the largest negative patch at pH 8.0.largest_positive_patch_density_8.0: Charge density of the largest positive patch at pH 8.0.largest_negative_patch_density_8.0: Charge density magnitude of the largest negative patch at pH 8.0.
all_charge_spatial_moment_magnitude_6.0: Whole-structure charge spatial moment magnitude at pH 6.0.surface_charge_spatial_moment_magnitude_6.0: Surface-residue charge spatial moment magnitude at pH 6.0.all_charge_spatial_moment_normalized_6.0: Size- and charge-normalized whole-structure spatial moment at pH 6.0.surface_charge_spatial_moment_normalized_6.0: Size- and charge-normalized surface spatial moment at pH 6.0.charge_separation_magnitude_6.0: Distance-weighted separation of positive and negative charge centers at pH 6.0.all_charge_spatial_moment_magnitude_7.0: Whole-structure charge spatial moment magnitude at pH 7.0.surface_charge_spatial_moment_magnitude_7.0: Surface-residue charge spatial moment magnitude at pH 7.0.all_charge_spatial_moment_normalized_7.0: Size- and charge-normalized whole-structure spatial moment at pH 7.0.surface_charge_spatial_moment_normalized_7.0: Size- and charge-normalized surface spatial moment at pH 7.0.charge_separation_magnitude_7.0: Distance-weighted separation of positive and negative charge centers at pH 7.0.all_charge_spatial_moment_magnitude_8.0: Whole-structure charge spatial moment magnitude at pH 8.0.surface_charge_spatial_moment_magnitude_8.0: Surface-residue charge spatial moment magnitude at pH 8.0.all_charge_spatial_moment_normalized_8.0: Size- and charge-normalized whole-structure spatial moment at pH 8.0.surface_charge_spatial_moment_normalized_8.0: Size- and charge-normalized surface spatial moment at pH 8.0.charge_separation_magnitude_8.0: Distance-weighted separation of positive and negative charge centers at pH 8.0.
nearest_opposite_charge_distance: Shortest anchor-point distance between oppositely charged residues at pH 7.0.nearest_positive_charge_distance: Shortest anchor-point distance between two positively charged residues at pH 7.0.nearest_negative_charge_distance: Shortest anchor-point distance between two negatively charged residues at pH 7.0.
pip install mcce-features
The CLI entry point will be available after a successful installation:
mcce-features
To extract features from an MCCE folder
mcce-features extract <folder_name>
To extract feature from multiple MCCE folders
mcce-features extract-folders <folder.lst>
The repo is not yet setup to publish automatically, so the usual pip install will be the original version.
# 1. clone this repo somewhere you want it to be, then enter its dir:
git clone https://github.com/pclass-lab/elefeatures.git
cd elefeatures
# 2. Activate an environment setup for python>-3.12, eg: p312:
conda activate p312
(p312)> pip install -e .
# 3. Go get features:
(p312)> cd my_mcce_simulations
# Extract features using the book.txt files from all the subfolders that have one:
(p312)> mcce-features extract-subfolders-with-book .
File <folder.lst> can be a (space, comma, tab) delimited file with a MCCE folder name in the first column of each line.