helmkit 0.1.0__tar.gz → 0.3.0__tar.gz

{helmkit-0.1.0 → helmkit-0.3.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: helmkit
-Version: 0.1.0
+Version: 0.3.0
 Summary: Parse HELM strings into RDKit molecules
 License-File: LICENSE
 Requires-Python: >=3.11
@@ -17,7 +17,7 @@ A Python library for converting HELM (Hierarchical Editing Language for Macromol
 from helmkit import Molecule
 
 # Create a molecule from a HELM string
-helm_string = "PEPTIDE1{A.R.G}$$$"
+helm_string = "PEPTIDE1{A.R.G}$$$$"
 molecule = Molecule(helm_string)
 
 # Access the RDKit molecule object
@@ -31,7 +31,7 @@ from helmkit import Molecule
 from rdkit.Chem import AllChem, Draw
 
 # Create a simple tripeptide (Ala-Arg-Gly)
-molecule = Molecule("PEPTIDE1{A.R.G}$$$")
+molecule = Molecule("PEPTIDE1{A.R.G}$$$$")
 
 # Generate 2D coordinates for visualization
 AllChem.Compute2DCoords(molecule.mol)
@@ -67,7 +67,7 @@ custom_sdf_path = "/path/to/your/custom_monomers.sdf"
 custom_monomers = load_monomer_library(custom_sdf_path)
 
 # Create molecule with custom monomer data
-molecule = Molecule("PEPTIDE1{A.R.G}$$$", monomer_df=custom_monomers)
+molecule = Molecule("PEPTIDE1{A.R.G}$$$$", monomer_df=custom_monomers)
 ```
 
 ## SDF File Structure Requirements
@@ -110,3 +110,22 @@ Ala
 
 $$$$
 ```
+
+## Parallel Processing of Peptides
+
+For workflows involving a large number of peptides, `helmkit` provides a function to process them in parallel, significantly improving performance.
+
+```python
+from helmkit import load_monomer_library
+from helmkit import load_peptides_in_parallel
+
+# Load your custom monomer data (optional)
+custom_sdf_path = "/path/to/your/custom_monomers.sdf"
+monomer_db = load_monomer_library(custom_sdf_path)
+
+# A list of HELM strings
+helm_strings = ["PEPTIDE1{A.R.G}$$$$", "PEPTIDE1{S.G.T}$$$$"]
+
+# Process peptides in parallel
+molecules = load_peptides_in_parallel(helm_strings, monomer_db)
+```

{helmkit-0.1.0 → helmkit-0.3.0}/README.md

@@ -8,7 +8,7 @@ A Python library for converting HELM (Hierarchical Editing Language for Macromol
 from helmkit import Molecule
 
 # Create a molecule from a HELM string
-helm_string = "PEPTIDE1{A.R.G}$$$"
+helm_string = "PEPTIDE1{A.R.G}$$$$"
 molecule = Molecule(helm_string)
 
 # Access the RDKit molecule object
@@ -22,7 +22,7 @@ from helmkit import Molecule
 from rdkit.Chem import AllChem, Draw
 
 # Create a simple tripeptide (Ala-Arg-Gly)
-molecule = Molecule("PEPTIDE1{A.R.G}$$$")
+molecule = Molecule("PEPTIDE1{A.R.G}$$$$")
 
 # Generate 2D coordinates for visualization
 AllChem.Compute2DCoords(molecule.mol)
@@ -58,7 +58,7 @@ custom_sdf_path = "/path/to/your/custom_monomers.sdf"
 custom_monomers = load_monomer_library(custom_sdf_path)
 
 # Create molecule with custom monomer data
-molecule = Molecule("PEPTIDE1{A.R.G}$$$", monomer_df=custom_monomers)
+molecule = Molecule("PEPTIDE1{A.R.G}$$$$", monomer_df=custom_monomers)
 ```
 
 ## SDF File Structure Requirements
@@ -101,3 +101,22 @@ Ala
 
 $$$$
 ```
+
+## Parallel Processing of Peptides
+
+For workflows involving a large number of peptides, `helmkit` provides a function to process them in parallel, significantly improving performance.
+
+```python
+from helmkit import load_monomer_library
+from helmkit import load_peptides_in_parallel
+
+# Load your custom monomer data (optional)
+custom_sdf_path = "/path/to/your/custom_monomers.sdf"
+monomer_db = load_monomer_library(custom_sdf_path)
+
+# A list of HELM strings
+helm_strings = ["PEPTIDE1{A.R.G}$$$$", "PEPTIDE1{S.G.T}$$$$"]
+
+# Process peptides in parallel
+molecules = load_peptides_in_parallel(helm_strings, monomer_db)
+```

helmkit-0.3.0/benchmarks/README.md

@@ -0,0 +1,55 @@
+# Benchmarks
+
+## Purpose
+
+This benchmark evaluates the **reading speed** of
+[`helmkit`](https://github.com/adaliaramon/helmkit) and
+[`pyPept`](https://github.com/Boehringer-Ingelheim/pyPept) for parsing HELM notation
+strings into RDKit molecules. The goal is to compare their performance and usability
+differences when processing peptides in HELM format.
+
+## Feature Comparison
+
+The libraries differ in key implementation aspects that affect speed and robustness:
+
+| Feature                       | `pyPept`                                             | `helmkit`                |
+| ----------------------------- | ---------------------------------------------------- | ------------------------ |
+| **Monomer library loading**   | Reloaded for every peptide                           | Loaded once per session  |
+| **Conversion path**           | HELM → BILN → Sequence → RDKit                       | HELM → RDKit             |
+| **Dependencies**              | `rdkit`, `pandas`, `biopython`, `requests`, `igraph` | `rdkit`                  |
+| **Monomer library input**     | Requires directory + library name                    | Requires only file path  |
+| **Directory format**          | Must be a Python module (with `__init__.py`)         | No such requirement      |
+| **Monomer format strictness** | Very strict (e.g., fails on parentheses, dashes)     | More tolerant            |
+| **Error handling**            | Exits the process on error                           | Raises Python exceptions |
+
+These differences influence both speed and ease of use, especially the monomer library
+loading and conversion path.
+
+## Reading Speed Results
+
+We used the [CycPeptMPDB](http://cycpeptmpdb.com/peptides/type_PAMPA/) dataset,
+comprising 7,298 cyclic peptides in HELM format. Peptides including monomers with names
+that `pyPept` cannot handle were excluded (monomers with names containing whitespace,
+hyphens or parenthesis), resulting in a total of 4694 peptides. We measured the total
+and average parsing time:
+
+| Tool                                | Total Time (s) | Avg Time per Peptide (s) |
+|-------------------------------------|----------------|--------------------------|
+| `pyPept`                            | 676.65         | 0.144                    |
+| `helmkit`                           | 3.60           | 0.00077                  |
+| `helmkit` (DB reload every peptide) | 261.81         | 0.056                    |
+
+`helmkit` outperforms `pyPept` by approximately 188× when loading the monomer library
+once per session. When forced to reload the library for every peptide (needs disabling
+memoization in the source code), `helmkit` is still about 3× faster.
+
+## Environment
+
+Benchmarks were run on an Intel Core i7-4790 (4 cores, 8 threads, 3.6 GHz) with 31.1 GiB
+RAM and SSD storage, using Python 3.12.10 on Arch Linux (kernel 6.15.7). Key package
+versions:
+
+- `polars 1.31.0` (CSV parsing and dataframe processing)
+- `rdkit 2025.3.3` (target output format and structure processing)
+- `pypept 1.0.0` (commit `ade9f5840691ad1f8fa22d13939a665c25175d5a`)
+- `helmkit 0.1.0` (local development version)