Table of Contents

• OpenAD
  • General
  • Workspaces
  • Toolkits
  • Runs
  • Utility
  • LLM
  • File System
  • Help
• DS4SD
  • Search Molecules
  • Search Collections
  • Collections
• RXN
  • Queries
  • Retrosynthesis
  • Prediction
• ST4SD
• GT4SD

OpenAD

See commands

General

openad Display the openad splash screen.

get status Display the currently selected workspace and toolkit.

display history Display the last 30 commands run in your current workspace.

clear sessions Clear any other sessions that may be running.

Workspaces

set workspace <workspace_name> Change the current workspace.

get workspace [ <workspace_name> ] Display details a workspace. When no workspace name is passed, details of your current workspace are displayed.

create workspace <workspace_name> [ description('<description>') on path '<path>' ] Create a new workspace with an optional description and path.

remove workspace <workspace_name> Remove a workspace from your registry. Note that this doesn’t remove the workspace’s directory.

list workspaces Lists all your workspaces.

Toolkits

ds4sd Display the splash screen for the DS4SD toolkit.

rxn Display the splash screen for the RXN toolkit.

st4sd Display the splash screen for the ST4SD toolkit.

gt4sd Display the splash screen for the GT4SD toolkit.

list toolkits List all installed toolkits. To see all available toolkits, run list all toolkits.

list all toolkits List all available toolkits.

add toolkit <toolkit_name> Install a toolkit.

remove toolkit <toolkit_name> Remove a toolkit from the registry.

Note: This doesn’t delete the toolkit code. If the toolkit is added again, a backup of the previous install is created in the toolkit directory at ~/.openad/toolkits.

set context <toolkit_name> [ reset ] Set your context to the chosen toolkit. By setting the context, the selected toolkit functions become available to you. The optional parameter reset can be used to reset your login information.

get context Display the currently selected toolkit.

unset context Exit your toolkit context. You will no longer have access to toolkit-specific functions.

Runs

create run Start recording a run.

save run as <run_name> Stop recording a run and save it.

run <run_name> Execute a previously recorded run. This will execute every command and continue regardless of any failures.

list runs List all runs saved in the current workspace.

display run <run_name> Display the commands stored in a certain run.

Utility

display data '<csv_filename>' Display data from a csv file.

-> result save [as '<csv_filename>'] Save table data to csv file.

-> result open Explore table data in the browser.

-> result edit Edit table data in the browser.

-> result copy Copy table data to clipboard, formatted for spreadheet.

-> result display Display the result in the CLI.

edit config '<json_config_file>' [ schema '<schema_file>'] Edit any JSON file in your workspace directly from the CLI. If a schema is specified, it will be used for validation and documentation.

show molecules using ( file '<mols_file>' | dataframe <dataframe> ) [ save as '<sdf_or_csv_filename>' | as molsobject ] Launch the molecule viewer to examine and select molecules from a SMILES sdf/csv dataset.

Examples:

• show molecules using file 'base_molecules.sdf' as molsobject
  
• show molecules using dataframe my_dataframe save as 'selection.sdf'
  
  

LLM

tell me <how to do xyz> Ask your AI assistant how to do anything in OpenAD.

set llm <language_model_name> Set the target language model name for the tell me command.

clear llm auth Clear the language model’s authentication file.

File System

list files List all files in your current workspace.

import from '<external_source_file>' to '<workspace_file>' Import a file from outside OpenAD into your current workspace.

export from '<workspace_file>' to '<external_file>' Export a file from your current workspace to anywhere on your hard drive.

copy file '<workspace_file>' to '<other_workspace_name>' Export a file from your current workspace to another workspace.

remove '<filename>' Remove a file from your current workspace.

Help

intro Display an introduction to the OpenAD CLI.

docs Open the documentation webpage.

? List all available commands.

<soft>...</soft> ? Display what a command does, or list all commands that contain this string.

DS4SD

See commands

Search Molecules

search for patents containing molecule '<smiles>' | '<inchi>' | '<inchi_key>' [ save as '<csv_filename>' ] Search for mentions of a specified molecules in registered patents. The queried molecule can be described as a SMILES string, InChI or InChiKey.

Use the save as clause to save the results as a csv file in your current workspace.

Example:
search for patents containing molecule 'CC(C)(c1ccccn1)C(CC(=O)O)Nc1nc(-c2c[nH]c3ncc(Cl)cc23)c(C#N)cc1F'

search for similar molecules to '<smiles>' [ save as '<csv_filename>' ] Search for molecules that are similar to the provided molecule or molecule substructure as provided in the <smiles_string>.

Use the save as clause to save the results as a csv file in your current workspace.

Example:
search for similar molecules to 'C1(C(=C)C([O-])C1C)=O'

search for molecules in patents from list ['<patent1>', '<patent2>', ...] | dataframe <dataframe_name> | file '<csv_filename>' [ save as '<csv_filename>' ] Search for molecules mentioned in a defined list of patents. When sourcing patents from a CSV or DataFrame, there must be column named “PATENT ID” or “patent id”.

Use the save as clause to save the results as a csv file in your current workspace.

Example:
search for molecules in patents from list ['CN108473493B','US20190023713A1']

search for substructure instances of '<smiles>' [ save as '<csv_filename>' ] Search for molecules by substructure, as defined by the <smiles_string>.

Use the save as clause to save the results as a csv file in your current workspace.

Example:
search for substructure instances of 'C1(C(=C)C([O-])C1C)=O' save as 'my_mol'

Search Collections

search collection '<collection_name_or_key>' for '<search_string>' [ using (page_size=<int> system_id=<system_id> edit_distance=<integer> display_first=<integer>) ] show (data | docs) [ estimate only | return as data | save as '<csv_filename>' ] Performs a document search of the Deep Search repository based on a given collection. The required using clause specifies the collection to search. Use estimate only to return only the potential number of hits.

Parameters:

• <collection_name_or_key> The name or index key for a collection. Use the command display all collections to list available collections.
  
• <search_string> The search string for the search.
  

The <search_string> supports elastic search string query syntax:

• + Signifies AND operation.
  
• | Signifies OR operation.
  
• - Negates a single token.
  
• \" Wraps a number of tokens to signify a phrase for searching.
  
• * At the end of a term -> signifies a prefix query
  
• ( & ) Signifies precedence
  
• ~N After a word -> signifies edit distance (fuzziness)
  
• ~N After a phrase -> signifies slop amount
  

Options for the using clause:

Note: The using clause requires all enclosed parameters to be defined in the same order as documented below.

• page_size=<integer> Result pagination, the default is None.
  
• system_id=<system_id> System cluster id, the default is ‘default’.
  
• edit_distance=<integer> Sets the search word span criteria for key words for document searches, the default is 5. When set to 0, no snippets will be be returned.
  
• display_first=<integer> When set, the displayed result set will be truncated at the given number.
  

Clauses:

• show (data | docs):
  
  • data Display structured data from within the documents.
    
  • docs Display document context.
    Both can be combined in a single command, e.g. show (data docs)
    
• estimate only Determine the potential number of hits.
  
• return as data For Notebook or API mode. Removes all styling from the Pandas DataFrame, ready for further processing.
  

Examples:

• Look for documents that contain discussions on power conversion efficiency:
  search collection 'arxiv-abstract' for 'ide(\"power conversion efficiency\" OR PCE) AND organ*' using ( edit_distance=20 system_id=default) show (docs)
  

• Search the PubChem archive for ‘Ibuprofen’ and display related molecules’ data:
  search collection 'pubchem' for 'Ibuprofen' show (data)
  

• Search for patents which mention a specific smiles molecule:
  search collection 'patent-uspto' for '\"smiles#ccc(coc(=o)cs)(c(=o)c(=o)cs)c(=o)c(=o)cs\"' show (data)
  
  

display collection matches for '<search_string>' [ save as '<csv_filename>' ] Search all collections for documents that contain a given Deep Search <search_string>. This is useful when narrowing down document collection(s) for subsequent search. You can use the <index_key> from the returned table in your next search.

Use the save as clause to save the results as a csv file in your current workspace.

Example:
display collection matches for 'Ibuprofen'

Collections

display collections for domain '<domain_name>' Display the available collections in a given Deep Search domain.

Use the command display all collections to find available domains.

Example:
display collections for domain 'Business Insights'

display all collections [ save as '<csv_filename>' ] Display all available collections in Deep Search.

Use the save as clause to save the results as a csv file in your current workspace.

display collections in domains from list <list_of_domains> [ save as '<csv_filename>' ] Display collections that belong to the listed domains.

Use the save as clause to save the results as a csv file in your current workspace.

Use the command display all collections to find available domains.

Example:
display collections in domains from list ['Scientific Literature']

display collection details '<collection_name>' | '<collection_key>' Display the details for a specified collection. You can specify a collection by its name or key.

Use the command display all collections to list available collections.

Example:
display collection details 'Patents from USPTO'

RXN

See commands

Queries

list rxn models Lists all RXN AI models currently available.

Retrosynthesis

predict retrosynthesis '<smiles>' [ using (option1=<value> option2=<value>) ] Perform a retrosynthesis route prediction on a molecule.

Options for the optional using clause:

• availability_pricing_threshold=<int> Maximum price in USD per g/ml of compounds. Default: no threshold.
  
• available_smiles='<smiles>.<smiles>.<smiles>' List of molecules available as precursors, delimited with a period.
  
• exclude_smiles='<smiles>.<smiles>.<smiles>' List of molecules to exlude from the set of precursors, delimited with a period.
  
• exclude_substructures='<smiles>.<smiles>.<smiles>' List of substructures to excludefrom the set of precursors, delimited with a period.
  
• exclude_target_molecule=<boolean> Excluded target molecule. The default is True
  
• fap=<float> Every retrosynthetic step is evaluated with the FAP, and is only retained when forward confidence is greater than the FAP value. The default is 0.6.
  
• max_steps=<int> The maximum number steps in the results. The default is 3.
  
• nbeams=<int> The maximum number of beams exploring the hypertree. The default is 10.
  
• pruning_steps=<int> The number of steps to prune a hypertree. The default is 2.
  
• ai_model='<model_name>' What model to use. Use the command list rxn models to list all available models. The default is ‘2020-07-01’.
  

Example:
predict retrosynthesis 'BrCCc1cccc2c(Br)c3ccccc3cc12' using (max_steps=3)

interpret recipe '<recipe_paragraph>' | '<txt_filename>' Build a ordered list of actions interpreted from a provided text-based recipe. The recipe can be provided as a string or as a text file from your current workspace.

Examples:

• interpret recipe 'my_recipe.txt'
  
• interpret recipe 'A solution of ((1S,2S)-1-{[(methoxymethyl-biphenyl-4-yl)-(2-pyridin-2-yl-cyclopropanecarbonyl)-amino]-methyl}-2-methyl-butyl)-carbamic acid tert-butyl ester (25 mg, 0.045 mmol) and dichloromethane (4 mL) was treated with a solution of HCl in dioxane (4 N, 0.5 mL) and the resulting reaction mixture was maintained at room temperature for 12 h. The reaction was then concentrated to dryness to afford (1R,2R)-2-pyridin-2-yl-cyclopropanecarboxylic acid ((2S,3S)-2-amino-3-methylpentyl)-(methoxymethyl-biphenyl-4-yl)-amide (18 mg, 95% yield) as a white solid.'
  
  

Prediction

predict reaction in batch from dataframe <dataframe_name> | file '<csv_filename>' | list '<smiles>.<smiles>' [ using (ai_model='<ai_model>') ] [ use_saved ] Run a batch of reaction predictions. The provided list of reactions can be specified as a DataFrame, a CSV file from your current workspace or a list of strings. When proving a DataFrame or CSV file, we will look for the “reactions” column.

Reactions are defined by combining two SMILES strings delimited by a period. For example: 'BrBr.c1ccc2cc3ccccc3cc2c1'

Options for the optional using clause:

• ai_model='<model_name>' What model to use. Use the command list rxn models to list all available models. The default is ‘2020-07-01’.
  

You can reuse previously generated results by appending the optional use_saved clause. This will reuse the results of a previously run command with the same parameters, if available.

Examples:

• predict reaction in batch from list ['BrBr.c1ccc2cc3ccccc3cc2c1CCO' , 'BrBr.c1ccc2cc3ccccc3cc2c1']
  
• predict reaction in batch from list ['BrBr.c1ccc2cc3ccccc3cc2c1CCO' , 'BrBr.c1ccc2cc3ccccc3cc2c1'] use_saved
  
  

predict reaction '<smiles>.<smiles>' [ using (ai_model='<ai_model>') ] [ use_saved ] Predict the reaction between two molecules.

Reactions are defined by combining two SMILES strings delimited by a period. For example: 'BrBr.c1ccc2cc3ccccc3cc2c1'

Options for the optional using clause:

• ai_model='<model_name>' What model to use. Use the command list rxn models to list all available models. The default is ‘2020-07-01’.
  

You can reuse previously generated results by appending the optional use_saved clause. This will reuse the results of a previously run command with the same parameters, if available.

Examples:

• predict reaction 'BrBr.c1ccc2cc3ccccc3cc2c1CCO'
  
• predict reaction 'BrBr.c1ccc2cc3ccccc3cc2c1CCO' use_saved
  
  

predict reaction topn in batch from dataframe <dataframe_name> | file '<csv_filename>' | list ['<smiles>.<smiles>','<smiles>.<smiles>'] [ using (topn=<integer> ai_model='<ai_model>') ] [ use_saved ] Run a batch of reaction predictions for topn. The provided list of reactions can be specified as a DataFrame, a CSV file from your current workspace or a list of strings. When proving a DataFrame or CSV file, we will look for the “reactions” column.

Reactions are defined by combining two SMILES strings delimited by a period. For example: 'BrBr.c1ccc2cc3ccccc3cc2c1'

Options for the optional using clause:

• ai_model='<model_name>' What model to use. Use the command list rxn models to list all available models. The default is ‘2020-07-01’.
  
• topn=<integer> Defined the number of results being returned. The default value is 3.
  

You can reuse previously generated results by appending the optional use_saved clause. This will reuse the results of a previously run command with the same parameters, if available.

Examples:

• predict reaction topn in batch from list ['BrBr.c1ccc2cc3ccccc3cc2c1CCO' , 'BrBr.c1ccc2cc3ccccc3cc2c1']
  
• predict reaction topn in batch from list ['BrBr.c1ccc2cc3ccccc3cc2c1CCO' , 'BrBr.c1ccc2cc3ccccc3cc2c1'] using (topn=6)
  
• predict reaction topn in batch from list ['BrBr.c1ccc2cc3ccccc3cc2c1CCO' , 'BrBr.c1ccc2cc3ccccc3cc2c1'] use_saved
  
  

ST4SD

See commands

GT4SD

See commands