A🍻e

Let the forcefield be your guide.

Believe me, I am a chemist....

[A🍻e @ade93]

I'm going to perish

[A🍻e @ade93]

Mak, anakmu gagal lagi, maafkan 🙏

[Diego del Alamo @DdelAlamo]

Time for another round of binder design

[A🍻e @ade93]

@pekerjalab Jilid buku Purcell 😲

[A🍻e @ade93]

Mimpi ngangenin orang itu kadang indah tapi kadang tai banget😭

[A🍻e @ade93]

Jujur suka banget sama sudut di foto ini, entah.

[A🍻e @ade93]

Jadi beberapa hari yg lalu nemu kamera murah, Sony a55, RP~1.5jt udah dapet lensa (yg tengah 18-55). Hari pergi ke Nakano buat nyari lensa, dapet 2 lensa murah ~130rb-an per buah. Yg satu OP banget sih buat harganya, Minolta ZOOM xi 35-200, pabrikan tahun 1991. Zoomnya motorize!

[A🍻e @ade93]

Pame lensa fullframe di kamera apsc itu emang ngebuat objeknya terlalu ke isolasi, jadinya susah ngambil suasana.

[ベイスターズ情報bot @bay_bot]

【デイリー】ＤｅＮＡ １１カード連続勝ち越しなし 決定打が出ず相川監督「なかなか苦しい結果」
https://t.co/jcWGBil2Ce #baystars

[Biology+AI Daily @BiologyAIDaily]

Ovo, an open-source ecosystem for de novo protein design

1 Ovo is presented as an end-to-end, infrastructure-agnostic platform that consolidates de novo protein design models, scalable workflows, data management, and interactive visualization—aiming to reduce the “glue work” typically needed to stitch together RFdiffusion/MPNN/AlphaFold-style pipelines.

2 The core idea is workflow standardization plus scalability: Ovo uses Nextflow orchestration to run the same design protocols on a laptop, HPC (e.g., SLURM/PBS), or cloud (e.g., AWS Batch), with batching/parallelism to support campaigns that may generate tens to hundreds of thousands of designs.

3 Ovo combines multiple access modes for different user profiles: a web UI (Streamlit) for step-by-step job submission and interactive filtering/visual inspection (Mol*), a CLI for power users, and a Python API for programmatic execution and downstream analysis in notebooks.

4 A key engineering contribution is the data layer: designs, jobs, parameters, and computed descriptors are stored in a standardized SQL-backed data model with stable unique identifiers (Pools/Rounds/Projects; design IDs like ovo_buv_0045_cycle01), enabling reproducibility, retrospective analysis, and cross-run tracking.

5 For scaffold (motif) design, Ovo provides a modular pipeline chaining RFdiffusion backbone generation, LigandMPNN/ProteinMPNN sequence design (optionally with side-chain packing), and orthogonal refolding/validation via AlphaFold2 initial guess and/or ESMFold, with UI-driven contig construction and preview runs to validate inputs before scaling.

6 The paper emphasizes practical filtering at scale: Ovo computes backbone plausibility (e.g., DSSP secondary structure fractions, radius of gyration), sequence/manufacturability descriptors (charge at pH 7.4/5.5, pI, entropy, aromaticity, composition), and refolding metrics (mean PAE, pLDDT, design RMSD, native motif RMSD) with configurable acceptance thresholds.

7 For binder design, Ovo integrates two complementary protocols: RFdiffusion-based binder generation with ProteinMPNN + PyRosetta FastRelax and AlphaFold2 multimer initial-guess validation; and BindCraft (AF2 multimer backprop “hallucination” + redesign + filters), highlighting tradeoffs in diversity, compute efficiency, and GPU memory.

8 Binder workflows add interface-centric metrics for triage: hotspot engagement via backbone contact counts, PyRosetta interface descriptors (ddG, buried contact surface, SAP aggregation propensity), and AF2-derived complex confidence (iPAE/ipAE, target-aligned binder RMSD), enabling multi-criteria ranking beyond raw AF2 confidence.

9 Ovo also supports binder diversification via partial RFdiffusion: controlled noising/denoising around an existing binder-target complex, optionally restricted to selected residues/segments, to explore local structure space for affinity/specificity optimization while keeping length/segment order fixed.

10 ProteinQC is introduced as a contextual QC module: it computes a broad descriptor panel (including ESM-1v likelihood, ESM-IF likelihood, ProteinSol, hydrophobic patch fraction, electrostatics/patch analysis via PEP-Patch and APBS, and shape descriptors like asphericity) and compares designs to reference distributions (e.g., PDB) to flag liabilities and support data-driven filtering and reporting.

11 The ecosystem is designed to be extensible via plugins: lightweight Python packages can add Nextflow pipelines, alternative methods for steps, or new UI pages; the paper demonstrates integration of external tooling (ProteinDJ pipeline) and a Promb humanness-evaluation plugin as examples of community-driven expansion.

💻Code: https://t.co/Hmlam8mlse
📜Paper: https://t.co/xbxNKldqju
#ProteinDesign #DeNovoProteinDesign #ComputationalBiology #Bioinformatics #Nextflow #AlphaFold2 #RFDiffusion #ProteinMPNN #OpenSource #MLOps

[A🍻e @ade93]

@kozirama Kuliahan zaman sekarang gak ada yg IPKnya kecil kang. Nu belet aja bisa diatas 3 😅

[A🍻e @ade93]

@afrkml Saya pernah hampir pingsan duakali. 1. Gara-gara salatri belum makan, terus disuruh lari keliling lapangan kemudian baris-berbaris dengan orang yg bau badan. Mual, lemes, pandangan berbintik2 gelap, hampir jatuh tapi ada yg nangkep. Gak jadi pingsan.

[Biology+AI Daily @BiologyAIDaily]

ProteomeLM: A proteome-scale language model enables accurate and rapid prediction of protein–protein interactions and gene essentiality across taxa

1. ProteomeLM is a transformer that operates on entire proteomes (not single proteins or local genomic windows), learning organism-level constraints that help predict protein–protein interactions (PPI) and gene essentiality across the tree of life.

2. Key idea: represent each protein by a pretrained sequence embedding (ESM-Cambrian, 1152D averaged over residues), then train ProteomeLM with a proteome-level masked objective: mask many proteins and reconstruct their embeddings using the remaining proteins in the same proteome.

3. A major design choice is dropping genomic positional encoding (problematic across taxa, especially eukaryotes) and instead adding a functional encoding derived from OrthoDB orthologous groups, sampled at multiple taxonomic levels to expose the model to coarse-to-fine functional identity during training.

4. Technical novelty in training: a custom “polar loss” for reconstructing continuous embeddings, designed to avoid a degenerate solution where the model collapses to reproducing the functional encoding; it separately penalizes residual direction (cosine term) and magnitude (squared norm difference).

5. Unsupervised PPI signal emerges in attention: without any interaction labels, specific attention heads strongly separate interacting vs noninteracting pairs across species (e.g., AUC up to ~0.92 in E. coli for a single head; strong cross-species heads also observed), with PPI signal enriched in intermediate layers.

6. The attention signal captures multiple relationship types: direct physical binding, same-complex membership, and broader genetic/functional associations (e.g., coexpression). It tends to recover broad functional associations and complex membership particularly well, and can partially disentangle physical vs functional relationships.

7. Interactome-wide screening: a lightweight logistic regression on ProteomeLM attention heads (trained with only ~100 positives + 1,000 random negatives) enables proteome-scale PPI ranking that is far faster than DCA pipelines and more accurate in large screens (e.g., human AUC ~0.83 vs ~0.73 reported for DCA; substantially higher recall among top-ranked pairs).

8. Compute advantage: once ProteomeLM is trained, inference (including ESM-C embeddings + attention extraction) is reported to take under ~10 minutes per proteome (e.g., human) on a single GPU, avoiding per-pair MSA building and per-pair model fitting required by DCA; the paper estimates up to ~6 orders of magnitude speedup for inference in human-scale screens.

9. Supervised PPI prediction (ProteomeLM-PPI): a downstream network that combines node features (ESM-C + ProteomeLM embeddings) and edge features (attention coefficients) reaches state-of-the-art or comparable performance across benchmarks, with especially strong cross-species generalization (notably large AUPR gains on E. coli when trained on human PPI).

10. Gene essentiality (ProteomeLM-Ess): a simple supervised classifier on ProteomeLM embeddings improves over using protein-only ESM-C embeddings, reaching strong performance across 91 species (OGEE) and generalizing to held-out proteomes and even minimal/synthetic cells (JCVI-Syn1.0, JCVI-Syn3A), with best reported AUC up to ~0.93 depending on model size/layer.

💻Code: https://t.co/Mdt8Y8pZaP
📜Paper: https://t.co/0JkNkHPGIx
#ComputationalBiology #Bioinformatics #ProteinInteractions #PPI #ProteinLanguageModels #Transformers #SystemsBiology #GeneEssentiality #FoundationModels #Proteomics

[Biology+AI Daily @BiologyAIDaily]

SurfDesign: Effective Protein Design on Molecular Surfaces

1. SurfDesign reframes protein design around molecular surfaces (shape + physicochemical complementarity), aiming to better control functional regions like binding interfaces and enzyme pockets compared with backbone-only conditioning.

2. Core technical idea: treat molecular surfaces as continuous geometric manifolds rather than unordered point clouds/meshes, so the model can leverage local tangent structure, curvature, and directional consistency that are important for interaction-centric design.

3. The method builds an oriented surface point cloud Q where each point includes coordinates, a unit normal vector, and physicochemical attributes (e.g., hydrophobicity, charge, H-bond features). Surfaces are generated via PyMOL/MSMS and denoised with Gaussian smoothing; no residue identities, MSAs, or functional labels are used in surface generation.

4. SurfDesign introduces a Surface-conditioned Equivariant Message Passing (SEMP) encoder: SE(3)-equivariant updates use invariant radial distances, curvature descriptors (from local covariance eigenvalues), and directional angular features derived from surface normals (two intersection angles + one dihedral angle).

5. Directionality is encoded with spherical Fourier–Bessel bases over distances and angles; messages are attention-reweighted and used to update both node features and coordinates, with optional per-layer recomputation of normals/curvatures to stay consistent as coordinates evolve.

6. To address limited surface-structure paired data and improve sequence priors, SurfDesign integrates pretrained protein language models (PLMs) via parameter-efficient fine-tuning (hybrid PEFT: structural adapter + LoRA), trained with conditional masked language modeling rather than autoregressive decoding.

7. Binder design benchmark (6 targets) uses AF2 pAE_interaction as a functional proxy. SurfDesign achieves the best average pAE_interaction (15.85) and the highest overall success rate (30.14%), outperforming SurfPro (surface-conditioned baseline) and backbone-only baselines like ProteinMPNN, PiFold, and LM-DESIGN.

8. Enzyme design benchmark (5 enzyme–substrate systems; leakage-controlled by excluding overlaps with CATH pretraining) uses ESP score as a proxy for enzyme–substrate compatibility. SurfDesign attains the best average success rate (47.30%) and the best average ESP under greedy decoding (0.9058), with gains persisting in a zero-shot substrate setting.

9. Inverse folding is positioned as a diagnostic for structural compatibility (not “recovering the native sequence”). SurfDesign reports strong results on CATH splits, including perplexity 2.41 and AAR 74.13% on CATH 4.2, plus improved surface recovery metrics (IoU/CD/NC) versus PLM-based baselines; scaling larger PLMs further improves recovery.

💻Code: https://t.co/xsjUXVHQVj
📜Paper: https://t.co/zMETTEXleB
#ProteinDesign #ComputationalBiology #GeometricDeepLearning #ProteinEngineering #EnzymeDesign #ProteinBinding #EquivariantNetworks #ProteinLanguageModels #KDD2026

[A🍻e @ade93]

@arieparikesit Skill isssue, terutama karena silver skinnya gak dikupas dulu. Jadi alot, terus enek pas ngunyah kelamaan. Gak enak deh. 😂

[A🍻e @ade93]

Hasrat ingin menghilang dari dunia mulai muncul setiap saat lagi😅

[A🍻e @ade93]

@sociotalker Setiap keluar dari lab langsung muntah-muntah🙏🙏🙏