Release Notes

Small Molecule Drug Discovery

Platform Environment

Maestro Graphical Interface

• Interactive 2D Overlay: Precision marquee and lasso selection on the 2D Overlay that synchronizes instantly to the Workspace, bypassing the visual clutter of the 3D binding site.
• Task Tool Smart Search: Intelligent, fuzzy search that understands user intent – instantly locating the desired panels even with typos or conceptual queries like ‘dockng’ or ‘T-cell’
• Maestro Assistant:
  • Agentic Status Indicators: Real-time status updates replace the generic loading animation, so you always know what the assistant is doing.
  • Multiple persistent chat sessions – Users can now create, rename, switch between, and delete multiple chat sessions, each maintaining its own conversation history.
  • Privacy & Transparency Controls: Users can now enable or disable the Maestro Assistant directly from Preferences, with clear in-UI communication of the current privacy mode. The updated welcome screen and persistent AI disclaimer footer reinforce that your workspace structures are never sent to the LLM.
• Surfaces: Projecting ESP and Fukui reactivity maps onto Jaguar-computed electron density isosurfaces is now a single-click action, where previously it required multiple manual steps.

Force Field

• Improvement in the accuracy of metal complexes and metal-ligand interactions of specific metal ions (Ca(II), Mg(II), Fe(II), Zn(II), Pt(II), Ir(III)) from the addition of the metals polarizable force field via OPLS5. Current application support is in relative/absolute binding FEP+ and Desmond (beta).

Target Validation & Structure Enablement

Protein Preparation

• Improved ease of use to assess clashes with crystal mates in the Diagnostics panel
• New and improved side-chain reconstruction algorithm
• Better logging of use and matching of FASTA sequences to entry
• Academic Maestro can now minimize prepared structures
• Automatically fix bonding of sulfoximines, which are often incorrectly assigned in the wwPDB/CCD 
• New option to automatically add glycosylation bonds in Maestro 
• Detect ring spears during preprocessing, displaying them in the ‘Overlapping’ Diagnostic pane

Predictive Tox Panel

• Full release of the Predictive Tox panel for screening unknown liabilities
• Open beta release of the Predictive Tox SAR panel for automated production of high accuracy atomistic models of common anti-targets using existing SAR

Binding Site & Structure Analysis

Mixed Solvent MD (MxMD)

• Save a manual step with automatic execution of SiteMap rescoring as part of MxMD simulations
• Density maps are now written in ccp4 format instead of cns

Hit Identification & Virtual Screening

Docking

• The high-accuracy MacroDock macrocycle docking workflow is now available from the command line
• Enhanced performance and efficiency in ultra-large library screening with Generative Glide (Beta): Efficiently screen ultra-large on-demand synthesizable libraries with generative AI in a fraction of the time as AL-Glide and brute force approaches

Ligand Preparation

Enumeration

• New interface to enumerate novel molecules using generative AI for molecular materials with the REINVENT method

Shape Screening

• QuickShape now stores oned_screen results as .csv

ABFEP

• Apply restraints from the interface

Lead Optimization

FEP+

• Identify the optimal simulation time to maintain accuracy with time-sliced statistics
• Share FEP+ protocols across Web Services (GraphDB), Active Learning FEP, and the FEP+ Protocol Builder for consistency with easy import and export of new FEP+ protocol configuration files
• Improved Force Field parameter merging
• Improved cycle-closure calculation performance by up to 100x by saving subcycles and graph traversal

Protein FEP

• Renamed protein “selectivity” to “affinity” to be consistent with residue scanning

FEP+ Protocol Builder

• Account for “grouped” compounds when assigning test/train splits

Macrocycles

• Macrocycle Score: New semi-automated workflow to optimize and score macrocyclic cores for a given set of conformers, with support for high-fidelity scoring with MLFFs and QM
• Macrocycle Stability: New script for using binding pose metadynamics to rank macrocyclizations for stability compared to a reference molecule
• Macrocyclize: More detailed rejection reasons are now reported for cyclization reactions
• Macrocyclize: Automated enumeration of all L- and D-amino acid attachment pairs for custom SMILES linkers in peptide mode
• Prime Macrocycle Sampling: New command-line argument, -use_random_seed, to set a random seed during sampling
• Prime Macrocycle Sampling: New command-line argument, -curvature_tol, to specify curvature restraints during sampling
• Prime Macrocycle Sampling: New command-line argument, -freeze_torsions, to specify torsions to freeze during sampling

De Novo Design

AutoDesigner – R-group Design

• Full release of R-group Design panel
• Improved AutoDesigner PDF report

AutoDesigner – Core Design

• Full release of Core Design panel
• Improved AutoDesigner PDF report

AutoDesigner – Linker Design

• Full release of Linker Design panel
• Improved AutoDesigner PDF report

Drug Formulations

• Formulation ML: Support for proteins

Crystal Structure Prediction

• Full release of Crystal Structure Prediction with support of anhydrous Z’=2 and monohydrate crystal polymorph structures: Identify stable crystal polymorphs at zero Kelvin and room temperature for a given compound through hierarchical scoring

Docs Content

• Filter tutorials based on interest
• Materials Science Panel Explorer URLs retain filter states
• New interactive GPU licensing calculator
• Collapsible sidebar menu

Education Content

• New Tutorial: Filtering and Validating Protein-Protein Docked Poses with Macromolecular Pose Filtering and MM-GBSA Residue Scanning
• New Tutorial: Introduction to Performing Metadynamics Simulations with Desmond
• New Tutorial: Screening Ultra-large Libraries with Generative Glide
• New Tutorial: Peptide Cyclization
• New Tutorial: Preparing Cyclic Peptides and Aligning them to a Reference Structure
• New Interactive Tutorials: Three interactive tutorials for AutoDesigner, a de novo design algorithm for rapidly exploring large chemical space for lead optimization
  • Set up Linker Design with AutoDesigner in Maestro 
  • Set up Core Design with AutoDesigner in Maestro
  • Set up R-Group Design with AutoDesigner in Maestro

Biologics Drug Discovery

• Non-standard amino acids script (Open Beta). A command line script to simplify and speed creation of non-standard amino acid databases for residue scanning.
• Introduced a “Fast Mode” command line option for calculating protein descriptors. The new mode speeds up computation of these descriptors by about 20% on a benchmark dataset. Especially useful for large ensembles of structures.
• Enabled light-chain template subtype selection during antibody modeling. This provides greater flexibility when selecting templates for generating antibody structural models.
• Added the lengths of antibody Complementarity-Determining Regions (CDRs) to the output of calculated protein descriptors for antibody profiling analyses.

GUI for Quantum ESPRESSO

Product: Quantum ESPRESSO (QE) Interface

• Reduced disk space usage for calculating partial charges
• Bader charge analysis (command line, requires Badar module installation)
• Parallelization over both bands and tasks (command line)
• Support for MLFF in NEB calculations

Defect Properties:

Product: MS DefectPro

• Defect Formation Energy Diagram: Analysis panel for defect energy

Microkinetics

Product: MS Microkinetics

• Improved UI for renaming reactions
• User control of displayed stage in the Final State tab

Active Learning Optoelectronics

Product: Active Learning Optoelectronics

• (+MATSCI_OPTO_AL_ML) Access to ML property prediction models
• Prediction for fluorescence

Reactivity

Product: MS Reactivity

• Reaction Network Viewer: Color scheme to identify reactants
• Nanoreactor: Setup for number of stored structures to calculate biasing potential (command line)
• Nanoreactor: Improved sampling of input conformations
• Nanoreactor: Option to use AutoTS for transition state analysis

Crystal Structure Prediction

Product: Crystal Structure Prediction

• Crystal Structure Prediction: Support for Z’=2 and monohydrates

Advanced Force Field Applications

Product: MS FF Applications

• Infrastructure: Support for metals with OPLS5 from the materials science workflows (open beta)

Transport Calculations via MD simulations

Product: MS Transport

• Diffusion Coefficient: Support for multi-component diffusion analysis
• Ionic Conductivity: Output report in the driver log for command-line access
• Thin Plane Shear: Support for viscosity prediction from velocity profile

Coarse-Grained (CG) Molecular Dynamics

Product: MS CG

• Coarse-Grained Mapping: Standardized color scheme for Martini protein residues
• Backmapping: (+COARSE_GRAIN_BACKMAPPING) Algorithm to map coarse-grained structures to all-atom models
• Coarse-Grained FF Assignment: Automatic setup of restraints for DPD mapping
• CG FF Builder: (+CGFF_BUILDER_PARAM_NO_RESPONSE) Improvements to fitting algorithms for bond and angle terms
• CG FF Builder: Martini type guesses made visible to users
• Particle and residue names visible in MS Maestro
• Improved Martini parameters for linear polysaccharides

Complex Bilayer Builder

Product: MS Complex Bilayer Builder

• Membrane Analysis: Automated head/tail assignment independent of pdb names

Formulation ML

Product: MS Formulation ML

• Formulation ML: Support for protein
• Formulation ML: Support for aggregation functions over ingredient descriptors

Layered Device ML

Product: MS Layered Device ML

• Support for using existing device models as descriptors for new device models

Generative AI

Product: MS Generative AI

• REINVENT: Generative AI for molecular materials using the REINVENT method

MS Maestro Builders and Tools

• Disordered System: Support for barrier potential with planar interface substrates
• Interface Builder: (+INTERFACE_BUILDER) Advanced solution for building interface models
• Materials Project: API key from materialsproject.org hidden from the panel UI
• Meta Workflows: Support for order parameter analysis
• Meta Workflows: Support for bulk macromolecule relaxation protocol
• Periodic Structure Enumeration: Option to rank by electrostatic energy
• Polymer: Speed-up for building ladder polymers
• Remove Molecules from System: Speed up of up to 100X for loading CG systems
• MD Multistage: Improvements to built-in relaxation protocol ‘Semicrystalline 1’
• Solvate System: Option to prevent molecular splits across periodic boundaries
• Sugar Builder: Stand-alone model building solution for common polysaccharides

Classical Mechanics

• Droplet: Coarse-grained structures enabled
• Electrolyte Analysis: Support for 2D distribution analysis of coordinating shells
• Electrolyte Analysis: Support for cluster composition analysis
• Electrolyte Analysis: Support for residence time analysis
• Electrolyte Analysis: A tab in the viewer panel to summarize analysis results
• Electrolyte Analysis: Display of RDF and structure factor plots
• Polymer Crosslink: Support for NpγT and NpAT ensembles
• Polymer Crosslink: Support for opening multiple viewer windows
• Surface Tension: (+SURFACE_TENSION_CG) Support for coarse-grained models
• Thermophysical Properties: Option to stop after a series of failed simulations
• Thermophysical Properties: Support for varying MD timesteps by temperature
• Thermophysical Properties: Display of Calibrated Tg value from the viewer panel
• Thermophysical Properties: Support for MLFF
• Umbrella Sampling: Option to export potential of mean force to CSV output

Quantum Mechanics

• Adsorption Energy: Atomic constraints setup for molecular QM calculations
• Adsorption Energy: Improved interface for setting atomic constraints
• Optoelectronic Film Properties: Option to select the excited states of interest for ISC/RISC
• Optoelectronic Film Properties: Improved parallelization for ISC/RISC
• Optoelectronic Film Properties: Geometry optimization settings for ISC/RISC

Docs Content

• Filter tutorials based on interest
• Materials Science Panel Explorer URLs retain filter states
• New interactive GPU licensing calculator
• Collapsible sidebar menu

Education Content

• New Tutorial: Protein Characterization: Part 1
• New Tutorial: De Novo Design of Novel Compounds with REINVENT
• New Tutorial: Machine Learning for Formulation Containing Proteins
• Updated Tutorial: Droplet Contact Analysis
• Updated Tutorial: Nanoreactor
• Updated Tutorial: Liquid Electrolyte Properties: Part 2
• Updated Tutorial: Defect Formation Energy Calculation
• Updated Tutorial: Crystal Structure Prediction
• Quick Reference Sheet: Sugar Builder

Education Content

Life Science

• New Tutorial: Filtering and Validating Protein-Protein Docked Poses with Macromolecular Pose Filtering and MM-GBSA Residue Scanning
• New Tutorial: Introduction to Performing Metadynamics Simulations with Desmond
• New Tutorial: Screening Ultra-large Libraries with Generative Glide
• New Tutorial: Peptide Cyclization
• New Tutorial: Preparing Cyclic Peptides and Aligning them to a Reference Structure
• New Interactive Tutorials: Three interactive tutorials for AutoDesigner, a de novo design algorithm for rapidly exploring large chemical space for lead optimization
  • Set up Linker Design with AutoDesigner in Maestro 
  • Set up Core Design with AutoDesigner in Maestro
  • Set up R-Group Design with AutoDesigner in Maestro

Materials Science

• New Tutorial: Protein Characterization: Part 1
• New Tutorial: De Novo Design of Novel Compounds with REINVENT
• New Tutorial: Machine Learning for Formulation Containing Proteins
• Updated Tutorial: Droplet Contact Analysis
• Updated Tutorial: Nanoreactor
• Updated Tutorial: Liquid Electrolyte Properties: Part 2
• Updated Tutorial: Defect Formation Energy Calculation
• Updated Tutorial: Crystal Structure Prediction
• Quick Reference Sheet: Sugar Builder

LiveDesign

What’s New in 2026-2

• Project-Level Scaffolds: Project admins can publish scaffolds from the R-group decomposition panel as a project-wide template, which users can import into any LiveReport’s R-group decomposition.

• Customizable Project Dashboards: customize the Project Dashboard Overview page by adding and arranging configurable widgets—including LiveReport Bookmarks, Entities, Resources, Comments, LiveReports, Templates, Project Milestones, and Plots

• (Beta) Multi-Entity Matrix Forms Widget: add a Multi-Entity Matrix widget to Forms layouts to view compound data in a transposed table, with compounds as column headers and assay, R-group, and model data as rows. Coordinate with your Schrodinger representative to enable this feature.

• Biologics

  • Toggle between atomistic and monomeristic rendering for biologic entities with up to 200 heavy atoms from the first-column entity image.

  • Design single-point mutants using custom monomers: custom and unnatural amino acid monomers now appear in the Mutations dropdown for residue substitution.

  • View and manage your Monomer Database with role-based access: Project Admins and Admins can create, edit, and delete monomers, while other users have view-only access; use the new list_monomers method in LDClient to query and filter monomers programmatically.

  • Search within annotated antibody regions: the subsequence search query now supports structure annotation, allowing searches within auto-annotated regions of an antibody; pair with an Entity type query to narrow the search scope.

  • Filter the Sequence Viewer by Framework regions alongside CDRs to support humanization, germline comparison, and liability assessment workflows.

  • Show all linkages in the Sequence Viewer to display covalent connection points for biologic sequences, including cyclic and branched peptides.

  • View logo plots in the Sequence Viewer to visualize amino acid conservation across aligned biologic sequences, with residues colored by their coloring scheme.

  • Residue sync selection between the 3D Visualizer and Sequence Viewer now works for structures uploaded as GE entities.

• 3D Visualization

  • Align structures in the 3D Visualizer by selecting a reference structure and aligning one or more mobile structures to it.

  • Pick custom colors for entire 3D structures in the Contents pane, available for GE structures and structures uploaded via Image/File Freeform columns.

• Advanced Search

  • Search by preferred alias using the new ID column query in the Advanced Search panel; ID queries can be combined with other query types, used within search groups, and set to Match by Child or Match by Property.

  • The Advanced Search complex view now includes Real/Virtual quick filter options and an Entity type dropdown for filtering by compounds, R-groups, biologics, and other entity types.

• RetroSynth

  • Improved performance with expanded building block coverage (170 billion building blocks) and search and route ranking improvements.

  • Simplified model paradigm: RetroSynth now uses a single protocol per instance rather than one per project; models remain project-specific. Note: all existing RetroSynth models have been archived as part of this change — add the RetroSynth model again to any LiveReports where it was previously configured. Requires LiveDesign ML.

• Plots: Aggregate Dates: group date-axis data by days, weeks, months, quarters, or years, and aggregate Y-axis values by Mean, Median, Count, or Sum; optionally show empty date bins for consistent axis spacing.

• Alias Model and Other Columns: assign custom display names to model columns and columns in the “Other” section of the Data & Columns tree at the project level (by project admins) or at the LiveReport level (by any user); aliases appear in column headers, filters, plot axes, and the Data & Columns tree, with the original system name visible in the column header tooltip.

• UX Improvements

  • Admin users can download a sample CSV template from the Bulk Add Users page to guide user upload file preparation.

  • The email field is now optional on the Update Users page in the Admin Panel.

  • The SSO login page now also displays username and password fields, allowing users to log in with standard credentials when SSO is unavailable.

What’s Been Fixed

• Advanced Search

  • Pasting IDs into the All IDs filter query box in complex view would also paste text outside the box and produce incorrect filtering results, and now IDs are pasted only within the filter query box.

  • Creating a new query group from an independent query in Advanced Search complex view would insert it out of order, and now appends the new group in the correct position.

  • Auto-update search for biologics would not include newly imported entities until the search was manually rerun, and now fetches matching entities as they are imported.

  • Toggling AND/OR logic between conditions in an Advanced Search query group would not switch to complex view as expected, and now correctly transitions to complex view when mixed logic is introduced.

  • New queries added in Advanced Search complex view would appear at a random position rather than at the cursor location, and now are inserted at the cursor position.

  • Creating a new query group via the cog menu in Advanced Search complex view would insert it at a random position, and now adds it in the correct location relative to the source query.

  • Toggling AND/OR logic between sub-queries within a query group in Advanced Search simple view would not apply the change, or would update all conditions simultaneously, and now correctly updates only the selected logic.

  • Deleting query groups containing subsequence queries in Advanced Search would throw an error that persisted even after a browser refresh, and now deletes without error.

  • Flipping AND/OR logic inside a query group with only 2 queries would unexpectedly switch the view from simple to complex, and now stays in simple view.

  • Using parentheses to group conditions inside a query group in Advanced Search complex view would always show an Invalid Search error, and now correctly allows valid groupings to be searched.

  • Advanced Search queries using date Freeform columns would fail to return results, and now searches correctly.

  • The cancel button now appears when casting a new vote.

  • Adding a plot to the Project Dashboard overview page would sometimes fail on the first attempt, and now adds reliably.

• Filters

  • Switching from complex to simple filter view would incorrectly warn that the filter expression was too complex in some cases, and now shows that warning only when the filter genuinely cannot be represented in simple view.

  • The filter view would automatically switch from complex to simple when a column with an active filter condition was removed from the LiveReport, and now retains the complex view.

• Formulas

  • Formulas using if() statements would fail to calculate when the if() statement used experimental assay cells that contained multiple values, in which one of those values was Null. Those formulas now calculate correctly.

  • Formulas using date functions on columns with multiple values per compound would cause all cells in the LiveReport to blink and become unusable, and now calculate correctly.

  • Formulas referencing an aggregated assay column would show ERROR instead of calculating, and now compute correctly for all aggregation modes.

  • Formulas using assay inputs whose first value was null would return ERROR instead of a result, and now calculate correctly.

  • The aggregateMax() and aggregateMin() formula functions now return blank cells instead of ERROR when the input column contains no values for a given compound.

  • The formula editor would show ‘Column not in LiveReport’ instead of the actual column name when a dependent column had been removed from the LiveReport, and now displays the column name correctly.

• Data & Columns Tree

  • The Data & Columns tree would sometimes show only certain sections after typing and clearing a search term, and now correctly resets to show all sections when the search is cleared.

• LiveReport

  • Duplicating, copying to a project, or saving a LiveReport as a template would fail when Advanced Search contained a subsequence query inside a query group, and now completes successfully.

  • Opening a LiveReport would load data for all plots in that LiveReport, causing browser performance issues, and now loads only the data needed for the active view.

  • Columns with long header text would not be resizable when a scroll bar appeared in the header, and now column width and height can be adjusted regardless of whether a scroll bar is present.

  • LiveReport subscription email notifications would not be delivered to users who authenticate via SSO, and now correctly send email notifications to SSO users.

  • Applying a template to a LiveReport that had previously had a template applied would silently fail with no changes made, and now applies correctly regardless of prior apply template history.

  • SVG images in Freeform columns and model result cells would display as blank cells instead of the image, and now display correctly across Chrome, Firefox, and Edge.

  • Column group names in the LiveReport header are now correctly centered.

  • Freezing or pinning a row would cause cells in that row to blink indefinitely until the browser was refreshed, and now rows freeze and pin correctly without persistent blinking.

  • The indicator showing that comments exist in a different mode (row-per-experiment or row-per-lot) was missing from comment column cells, and now appears correctly when switching between modes.

  • Switching to Kanban view would cause tiles to display flashing cells instead of data, and now correctly loads results for LiveReports with fewer than 1000 compounds.

  • Child entities with multiple lots created via the Entity Grouping Manager would continuously shift position in Row-per-Lot mode, and now remain stable.

  • Limited assay columns with filtered-out values would display alignment lines without proper spacing, breaking intracell alignment for multi-value cells, and now display with correct spacing.

  • Users assigned as editors to a read-only comment column could not add or edit comments, and now have the expected edit access in those columns.

• Maestro

  • Column names containing < or > characters would be truncated at those characters when importing data into Maestro, and now the full column name is preserved.

• Admin Panel

  • Atom font size and bond line width options are now available again in the Admin Panel Properties page.

  • Users created via bulk upload would not be assigned the license status specified in the CSV file, and now correctly inherit license and role settings from the uploaded file.

  • Assigning a license to an SSO user migrated from LDAP would fail if the user had no email address on file, and now succeeds with a notification displayed when no email is available.

  • Users created via bulk upload would lose their assigned roles if any field was edited before saving, and now retain their roles regardless of edits made during the import.

• Authentication

  • Logging in via SSO would intermittently show an error page in some configurations, and now correctly completes the SSO login flow.

  • Logging in via SSO would sometimes trigger an immediate logout followed by a repeated login/logout loop, and now completes login without looping.

  • Users would be logged out due to the idle timeout even while actively using LiveDesign in another browser tab, and now activity in any tab resets the idle timer across all open tabs.

  • Users would be logged out prematurely when the browser suspended an idle tab due to a network interruption, and now remain logged in through brief network disruptions.

• Biologics & Sketcher

  • Some toolbar buttons in the Sketcher would display in dark mode while others did not, and now all toolbar buttons display with a consistent appearance.

  • Structure images generated as SVGs would display an inconsistent font that did not match the Sketcher, and now correctly use the expected font.

• 3D Visualization

  • Lasso selection in the 3D Visualizer would select residues and atoms outside the drawn area, and now correctly selects only the residues and atoms within the lasso area.