Add: TENCoM vibrational entropy differential standalone tool#46
Open
lmorency wants to merge 2 commits intoclaude/write-implementation-MglRZfrom
Open
Add: TENCoM vibrational entropy differential standalone tool#46lmorency wants to merge 2 commits intoclaude/write-implementation-MglRZfrom
lmorency wants to merge 2 commits intoclaude/write-implementation-MglRZfrom
Conversation
Standalone executable (tencom_entropy_diff) for computing vibrational entropy differentials between a reference structure and multiple target structures using the Torsional Elastic Network Contact Model (TENCoM). Key changes: - Add TorsionalENM::build_from_ca() for standalone Cα-coordinate input without dependency on FlexAID atom/resid data structures - Add tencom_entropy_diff.cpp: standalone tool that parses PDB files, computes TENCoM normal modes, and outputs FlexibilityMode reports (ligand-free BindingMode analog for global flexibility analysis) - Outputs eigenvalue differentials, eigenvector overlaps, per-residue B-factor differentials, and vibrational entropy/free energy deltas - Full flexibility always ON by default - CMake option: -DENABLE_TENCOM_ENTROPY_DIFF=ON - 9 GoogleTest unit tests (all passing) https://claude.ai/code/session_01PGgbrgbfdcHiSBMLd8GRWB
…tal) Major performance acceleration of the Torsional Elastic Network Contact Model across all hot paths: CPU acceleration (tencm.cpp): - Eigen SelfAdjointEigenSolver replaces hand-rolled Jacobi diagonalisation (divide-and-conquer vs O(M³) per-sweep Jacobi — order-of-magnitude speedup) - Eigen Map<VectorXd> for sample() mode accumulation and strain energy (BLAS) - OpenMP parallel contact discovery with thread-local vectors - OpenMP parallel Hessian assembly with thread-local matrices + reduction - OpenMP parallel B-factor computation (independent per residue) - AVX-512 16-wide distance batching in contact discovery (with mask extract) - AVX-512 8-wide double FMA accumulation in B-factor inner loop - AVX2 8-wide distance batching via simd::distance2_1x8 fallback GPU acceleration (new files): - tencm_cuda.cu/cuh: CUDA kernels for contact discovery (one thread per pair, atomic compaction) and Hessian assembly (one block per contact, 256 threads handle (k,l) pairs with atomicAdd) - tencm_metal.mm/h: Metal compute shaders matching CUDA kernels, using CAS-loop atomic float add (Metal 2 compatible), embedded MSL source Multi-structure parallelism (tencom_entropy_diff.cpp): - OpenMP parallel for over target structures with mutex-guarded console I/O - std::atomic counters for success/fail tracking CMake integration: - CUDA/Metal sources added to FlexAID main target and tencom_entropy_diff - GPU_THRESHOLD = 256 residues (below that, CPU is faster due to PCIe cost) All 9 TENCoM entropy diff tests + all 42 statmech tests pass. https://claude.ai/code/session_01PGgbrgbfdcHiSBMLd8GRWB
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Standalone executable (tencom_entropy_diff) for computing vibrational
entropy differentials between a reference structure and multiple target
structures using the Torsional Elastic Network Contact Model (TENCoM).
Key changes:
without dependency on FlexAID atom/resid data structures
computes TENCoM normal modes, and outputs FlexibilityMode reports
(ligand-free BindingMode analog for global flexibility analysis)
B-factor differentials, and vibrational entropy/free energy deltas
https://claude.ai/code/session_01PGgbrgbfdcHiSBMLd8GRWB