1 Signing up

To sign up for the system, follow the link https://app.syntelly.com/ and click the "Sign up" button.

On the page that appears, enter your email address, name, position and organization, create a password to log into the System and enter an invite token (you can get an invite token by writing to admin@syntelly.com):

To complete registration, click "Sign up".

2 Signing in

To sign in, follow the link https://app.syntelly.com/, enter your email address, password and click the “Sign in” button:

After signing in, the “Search” section is displayed:

In the lower right corner, when opening the pages, there appear pop-up tips for working with the System “Tot tips”. You can close the hint by clicking on the cross.

When you are sure you are familiar with the platform's logic and all hotkeys, you can disable “Hot tips” by clicking on your profile icon

In the menu on the left you can select a work section, a language (Russian or English), and open the user’s personal account. To hide the menu, click the button

Here, in the user profile, you can also change your password or log out of the system

In the lower right corner also, you can open the feedback form to send an email to the Syntelly technical support.

3 Section «Search»

This section provides access to the main database containing organic compounds, reactions, as well as links to publications on chemical topics.

To open the section, click "Search" in the left panel.

3.1 Search by structures (molecules)

Molecules can be searched in two ways:

After that, run the search by pressing Enter or clicking the search icon.

Additionally, you can set the following search parameters in the "Filters" panel (the panel is automatically displayed after the first search):

Match type — the filter is used to find structures according to the correspondence of the given molecule:
Exact match (searching for a molecule completely matching the specified structure),
Substructural search (searching for molecules containing the drawn fragment),
Similar structures (searching by molecular similarity);
Markush structures (searching not for an individual compound, but for a group of chemical compounds)

Hints:

3.1.1 Working with found molecules

Using the buttons on the right below the search bar, you can adjust the scale of the cards in the displayed list of found molecules:

To add a molecule to your own dataset click on the plus sign on the molecule card

If you want to edit the found molecule, click on this button to open it in the molecular editor.

For each card the following is displayed:

• ID of the molecule in the Syntelly database;
• Structural formula of the molecule;
• Safety parameters of the molecule in terms of toxicity ("Tox"), physicochemical ("Phys"), biological ("Bio") and environmental ("Eco") properties:
• Red - molecule with high hazard indicators;
• Green - molecule with low hazard indicators;
• Yellow - molecule with average hazard indicators;
• Grey - the parameters haven't been calculated yet (open the structure card for calculation)
• «Literature» — link to the list of documents in which this compound occurs;
• «Reactions» — link to reactions in which the particular compound is involved.

The System allows you to:

• quickly copy SMILES
• download the structure in MOL format
• transfer to other modules
• copy link to the card
• download the structure image in PNG format;

To to use a required option, click the button
in the upper right corner of the card and select the appropriate command:

A card with preview information is displayed:

To view a molecule on a new (separate) page, left-click on the molecule card. A new page with various information blocks will open:

The molecule structure can be saved as a .png file. Also, the molecule structure and its properties can be saved in a .pdf file. To copy the link to the molecule card, click "Copy link".

To view detailed information on a molecule, but not open it on a new page, click the button on the card:

3.1.2 Properties of molecules

All calculated and experimental properties of molecules are contained in the blocks described below. To open properties, click “Search” or “Datasets” in the left menu, select a molecule and left-click on it.

Please note that if the database contains experimental values for the requested molecule, the system displays exactly them, and a green “EXP”indicator is displayed next to the parameters. If there is no experimental data on the molecule, then predicted values are displayed. This applies to all information blocks in the molecule card.

3.1.2.1 Structural data

The block contains information about the structure of the molecule:

3.1.2.2 Synonyms

The block contains all the synonyms known to the Syntelly team, which the molecule is found in various chemical databases, including the CAS number

3.1.2.3 Biological activity

The block contains the prediction of the biological activity of the molecule. It provides models for calculating the effectiveness of inhibition of five cytochromes: CYP1A2, CYP2C19, CYP2C9, CYP2D6, CYP3A4.

Cytochromes P450 are key enzymes that catalyze the metabolism (oxidation) of xenobiotics in the body. The biological function of cytochromes P450 is to remove toxic compounds from the body. Cytochrome inhibitors can stimulate undesirable side effects and lead to the accumulation of toxic compounds in the body. The prediction of cytochrome P450 inhibition makes it possible to evaluate the metabolism of the studied compounds, their safety and possible interactions with other drugs.

When you hover the cursor over the cytochrome name, the system displays tooltips for the following parameters:

The block also predicts the following properties of the selected molecule:

3.1.2.4 Physicochemical properties

The block contains physicochemical properties of the selected molecule:

9. Retention time – retention time in a standard chromatographic system
Experimental conditions:

10. Refractive Index – refractive index

3.1.2.5 Toxicity

The block contains the prediction of the hazard and toxicity of the compound.

Toxicity indicators are divided into two groups:

1. Animal lethal dose models
2. Effect-based models

Within the first group, the following indicators are predicted for different animal species:

LD50 (median lethal dose) – the average lethal dose of a chemical substance that causes the death of 50% of animals when introduced into the body; expressed in milligrams of substance per kilogram of animal weight (mg/kg).

LDLo (lowest lethal dose) – the lowest lethal dose causing death in a given animal species. The dosage is also indicated per unit of body weight (mg/kg).

1. Mouse Oral LD₅₀
2. Mouse Intraperitoneal LD₅₀
3. Mouse Intramuscular LD₅₀
4. Mouse Intraperitoneal LD_Lo
5. Mouse Intravenous LD₅₀
6. Mouse Skin LD₅₀ — lethal dose when applied to the mouse skin;
7. Mouse Subcutaneous LD₅₀
8. Mouse Intramuscular LD₅₀
9. Rat Oral LD₅₀
10. Rat Oral LD_Lo
11. Rat Intravenous LD₅₀
12. Rat Intraperitoneal LD₅₀
13. Rat Skin LD₅₀
14. Rat Subcutaneous LD₅₀
15. Rat Intraperitoneal LD_Lo
16. Rabbit Oral LD₅₀
17. Rabbit Intravenous LD₅₀
18. Rabbit Intravenous LD_Lo
19. Rabbit Skin LD₅₀
20. Dog Oral LD₅₀
21. Dog Intravenous LD₅₀
22. Cat Intravenous LD₅₀
23. Quail Oral LD₅₀
24. Bird Wild Oral LD₅₀
25. Chicken Oral LD₅₀
26. Guinea Pig Oral LD₅₀
27. Dog Intravenous LD_Lo
28. Guinea Pig Intraperitoneal LD₅₀

The second group of predicted toxicity properties consists of the following parameters:

1. Selective Target-organ Or System Toxicity Single Exposure
2. Acute Toxicity Swallowed (GOST 56957-2016)
3. Eye Corrosion
4. Eye Irritation
5. Developmental Toxicity
6. Reproductive Toxicity
7. Cardiotoxicity
8. Hepatotoxicity
9. Carcinogenicity
10. DILI (Drug-Induced Liver Injury)
11. Ames test

3.1.2.6 Ecology properties

The block displays the environmental properties of compounds according to several

The System calculates the following parameters:

1. 40 hour Tetrahymena pyriformis IGC50 – concentration of the test substance in water in mg/l leading to 50% inhibition of the growth of Tetrahymena pyriformis after 40 hours.
2. Bioconcentration factor
3. 48 hour Daphnia magna LC50 – the model provides the quantitative prediction for Daphnia Magna LC50 (48 hours) expressed in -log(mol/L) and converted to mg/L.
4. «96 hour Fathead Minnow LC50» – concentration of the substance at which 50% of the population of Fathead Minnow (a species of freshwater fish) is expected to die in 96 hours. This is an important indicator of the toxicity of the substance to fish, and helps assess risks to aquatic ecosystems.
5. Acute Aquatic Toxicity (GOST 57455 of 2017)

3.1.2.7 Assessment of complexity of synthesis

This block provides an assessment of synthetic complexity based on various predictive models.

В системе рассчитываются следующие параметры:

1. «SYBA» – Bayesian estimation of synthetic accessibility of organic compounds.
2. «Complexity (SCScore: Synthetic Complexity Learned from a 12 million Reaction and predicted by neural network)» – Connor Coley scale, where “1” signifies that the compound is simply synthesized, and “5” signifies it is complex to synthesize.

3.1.2.8 Druglikeness

This block represents the parameters of similarity to drug molecules.

3.1.2.9 External databases

The block is displayed if there are links to external databases for the molecule!
The block contains a list of external databases in which the molecule occurs. You can use the links to go to the original source page.

3.2 Applicability of models

For models that predict biological, physicochemical, toxicological, and environmental properties, you can enable a display of the applicability level of the model to the specific molecule being viewed.

To display applicability of the model, turn on the switch in the upper right corner, on the molecule card.

When the switch is turned on, the percentage of applicability of the model to the molecule is displayed next to each property:

• If the indicator is from 0 to 20%, the prediction reliability is low. There were few molecules similar to the selected one in the model's training data
• 20-50% Average prediction reliability
• 50-100% High prediction reliability
• “Not applicable” the model cannot be applied to predict the properties of this molecule
• Without:

3.3 Search by reactions (in develop*)

In this section you can search for reactions by reactant, reagent or final reaction product.

To open the section, select “Search” in the menu on the left and click the “By reactions” tab.

Molecular structures (reagent, reactant or final product) can be searched in two ways:

• In the search bar, you can enter the name of the molecule in any convenient format: SMILES, synonyms, trade name, vendor code, name in IUPAC format, CAS number and code designations from other databases.

–OR–

• In the molecular editor draw the structural formula of the molecule. To do this, click the “Draw” button, draw a molecule in the window that appears and click “Save”.

Additionally, you can set the following search parameters in the “Filters” panel (the panel automatically appears after the first reaction search):

• «Match type» — the filter is used to find structures according to the correspondence of the given molecule. For example, you can search for an exact match to a structure or perform a substructural search.
• “Role of substance” — the filter is used to search for a reaction by the role of substances in the reaction, for example, you can search for reactions in which a certain compound is the product, reactant or reactant.
• "Yield, %" — the filter is used to search for a reaction with a certain percentage of product yield.

To remove all configured settings, click "Clear all filters".

As a result, the System displays a list of reactions that match the conditions of the request.
The reactions found can be sorted by relevance, publication date, reaction output, and number of steps.

When clicking on the reaction card, you can open it in a separate window and view detailed information:

3.4 Search by literature

In this section you can search through the scientific chemical literature collected in the Syntelly database. To open this section, select “Search” in the menu on the left, then open the “By literature” tab.

Please note!

1. If any molecule identifier (SMILES, IUPAC or CAS number) is entered into the search bar, a search for this compound will be carried out in the literature. Also, to search by structure in documents, the required compound can be specified in the molecular editor:

2. If text is entered into the search line, the corresponding text will be searched in the titles and abstracts of documents.

To initially specify the search query with restraining conditions, such as:

• Authors,
• DOI,
• Patent number
• Journal
• Applicant
• Claims
• Full text
• Publisher
• Owner

Click “Set up conditions” under the search bar

Please note!
A search by conditions can be performed, including if there is no query in the search line.

The search conditions can be combined. To do this, in the opening window, you can select conditions in any combination and specify the required values. To combine several conditions into one search query, you can use the logical operators And or OR. If And is selected, one of the conditions is used.

Click the "Confirm" button.

After completing the search query, you can refine it by selecting filters in the window that opens on the left.

After applying all the parameters, the System will adjust the list of found publications.
For any found article, you can copy the link, DOI (unique publication ID) or hide the article from the search results by clicking on "..." in the article card.

To view the full text of the publication, structures or reactions that are in the publication, click "Read full".
To view the substances mentioned in the document, click "Structures".
And to open reactions, click "Reactions".
The buttons are at the bottom of each card:

To view full information about the article, left-click on it.

• "Document type" — type of publication (journal article, patent);
• "Language" — language selection: Russian, English;
• “Publication date” — the date of publication of the document.

Please note!
If you have already specified the desired value for the year of publication earlier, when setting search conditions, then this selection field in the filters will not be available. To change it, you need to make adjustments specifically to the query constructor.

To remove all configured settings, click "Clear all filters". To apply filters, click "Apply".

Additionally, you can set the following search parameters in the "Filters" panel:

4 Section "Molecular Editor"

This section allows you to predict the properties of compounds that are not in the Syntelly database.

There are two ways to insert a molecule:

1. Enter the brand name, IUPAC name, supplier code or CAS number in the search bar and click .

You can display the structure of an already known molecule, make adjustments to it in accordance with your experiment and predict the properties of a new compound.

2. Enter SMILES by clicking the appropriate button.

Hint: To delete a text query, click the button

3. Draw a structure in the molecular editor.
To show the properties of the drawn molecule, click the button
in the upper right corner of the molecular editor page.

There will be displayed a card of the molecule with its calculated properties:

To export the molecular structure to a .png file, click “Download PNG”.
To export the molecule and all the calculated properties to a .pdf file, click

“Download PDF”.
And also you can Download MOL. If you want to add to your dataset click “To dataset”

To return to the molecular editor, click the button in the upper right corner of the page.
A detailed description of all calculated properties can be found in the section Calculated properties of molecules.

5 Section «Datasets»

This section is designed to work with the user's personal datasets, as well as with thematic datasets provided by the Syntelly team.

To open the section, click "Datasets" in the menu on the left, then select the "Thematic" or "Personal" tab.

5.1 Creating a dataset

While in the “Datasets” section, click “Private” then

Next, enter the name and description of the dataset and click “Create”:

After creating a dataset, it appears in the "Datasets" section in the "Private" tab.

5.2 Adding molecules to a dataset

To add a molecule to a dataset, in the menu on the left, select the “Datasets” section > “Thematic” or “Private” tab and click the button at the top of the page or, if the dataset has just been created, click the “Add Molecule” button in the middle of the page.

Next, you need to select the molecule uploading method from the drop-down list:

• «SMILES» — after selecting this option, enter SMILES in the field that appears;
• «Visual editor» — after selecting this option, the molecular editor, is displayed, where you can draw the required structure;
• «Upload from file» — after selecting this option, click the button
• "Select a file" and upload a file in .sdf, .csv, .smi format;

Please note!
The uploaded file must have a column containing SMILES of compounds. When downloading, you will be prompted to select the desired column by the column names of your file.

• «Search by synonyms» — after selecting this option, enter a synonym (the name of the molecule) in the field that appears.

The found molecule is displayed in the preview window. To add it to the dataset, click the "Upload" button:

5.3 Viewing molecule data in a personal dataset

To view molecules, left-click on the dataset. Working with molecules in a personal dataset is the same as in a public database.

5.4 Selecting a dataset

To select/deselect a dataset, click the button

Please note! You can also remove the selection by clicking on the right in an empty area of the screen.

and use the “Select” command:

5.5 Editing a dataset

To edit the name and description of the dataset, click the button

When editing, you can change the name and description of the dataset.

and select the “Edit” command:

5.6 Viewing log messages

To view log messages, you need to click the button

The log displays data on modifications of the dataset, modification dates and other information:

on the dataset and select the “Viewing log messages” command:

5.7 Deleting a dataset

Click the button

–OR–

Select a dataset and click the button

and select "Delete" in the menu that appears:

5.8 Copying a dataset

To copy a dataset with all molecules, select a dataset. To select a dataset, hold down the Ctrl key and click on the desired dataset (to select one at a time) or Shift (to select a group of datasets), then click the button

5.9 Merging datasets

To merge datasets into one, select two or more datasets and click the button

5.10 Showing a dataset on SynMap

To show a dataset on the SynMap coordinate plane, select a dataset and click the button

5.11 Exporting a dataset

To export a dataset with all molecules and their properties, select a dataset and click the button

If the properties of molecules are not calculated, the System asks whether it is necessary to calculate all the properties or export only SMILES:

Export can be done in .CSV or .SDF formats:

The file is downloaded to the local disk of the personal computer.

5.12 Calculation of molecule properties

To calculate the properties of compounds included in the selected dataset (toxicological, physicochemical, biological, etc.), select a dataset and click the button

5.13 Updating data

To update data, select a dataset and click the button

5.14 Undoing an action

To undo an action, make an action and click the button “Cancel” Not all actions can be undone

6 Section «SynMap»

This section allows you to visually project the structures of molecules on a two-dimensional plane so that molecules with similar properties appear nearby, forming clusters.

Additionally, you can generate chemical compounds with specified properties (QED, Boiling point, Melting point, Mouse oral LD50, LogP, LogS, DMSO Solubility, Brutto, MMap).

Hint:

• To be able to display a dataset on a 3D plane, its three-dimensional coordinates are calculated, so displaying large datasets can take considerable time.

6.1 Viewing groups of chemical compounds

This section allows you to get a quick and clear representation of the main groups of chemical compounds that are in the dataset. The model projects the structures of chemical compounds into the X and Y coordinates on a two-dimensional plane. The coordinates themselves do not carry any physical meaning, but the model is configured in such a way that structurally close compounds are located nearby together.
As a result, chemical compounds are distributed on the map so that you can see familiar clusters: a cluster of simple linear aliphatic compounds, a cluster of steroids, bisphenyls, psychoactive substances, etc.

This section can be opened in two ways:
1.In the "Datasets" section, select a prepared dataset and show it on SynMap.
-OR-
2.Open the "SynMap" section, and to display compounds on the map, click the "+" button in the "Layers" block.

Next, you need to select a pre-prepared public and/or personal dataset:

HINT: The System allows you to select several datasets at once.

After selecting the datasets, click “Select”.

Loading large datasets takes some time. If you want to make sure that it is in progress, pay attention to the "Layers" block in the upper right corner, the name of the dataset and the loading indicator will appear there. When the download is complete, you will see all molecules on the map.

Each point on this map is a molecule, and when you hover over the point, a card of this molecule is displayed. If several datasets are loaded at once, each of them is displayed in a separate color.
Using the map, you can easily determine which main classes of compounds predominate in this dataset. For example, in the lower part in the center there is a condensation of molecules with similar properties. The structure of each molecule can be viewed by hovering the cursor over it. In this case, simple molecules are located in this area:

You can select individual groups of molecules on the map, as well as export them to a separate dataset.

To select the desired area, click

or

and select, respectively, a rectangular or arbitrary area with the mouse.

Hint: To cancel a selection, double-click with the left mouse button outside the selected area.
At this stage, you can move the boundaries of the selected area. The maximum possible number of molecules for isolation and storage is 50,000.
A separate panel displays the structures that are included in the selected area:

By clicking on the button

Hints:

• The degree of transparency of the dots can be adjusted with the slider in the upper right corner.
• The map can be zoomed in or out by mouse scrolling or using the zoom buttons. In order to return to the original state of the map (default), click on the button (automatic scaling).
• To move the entire map using the mouse, click .

Molecules placed on the map can be displayed on a 3D plane. To do this, click on the corresponding button in the upper right corner of the map

The saved set can be found in the "Datasets" section.

Using the button

In the right area of the map we can see the class of steroid compounds.

, you can view information about the molecule, copy SMILES, open it in the molecular editor, download the structure in .png format, or add the molecule to another dataset. More detailed information on these actions is described in the section Working with molecular structures.

Selected molecules can also be saved in a new or already created dataset.
To save a structure to an existing dataset, click the “Save to dataset” button on the selected structure, select the set in the list and click the “Add to dataset” button:

To save the structure to a new dataset, enter the name of the new set, click the “Create new” button.

After the new dataset appears in the list of datasets, click on it, select it and click “Add to dataset”:

, you can turn on or off the display of a dataset on the map.

To delete a dataset from SynMap, click the button

6.2 Generating compounds with specified properties

To create new compounds, go to the “SynMap” section and click on the “+” in the “Generators” block.

Select the sector in which you want to create the generator:

• "Name" (arbitrary);
• Parent connection (for repeated generations);
• "Epochs" — the number of attempts by the system to come up with a new molecule; with an increase in the number of epochs, the neural network will have more time to generate compounds, this will indirectly affect the quality of generation (the more epochs, the more variants will be offered), but it is worth considering that the generation time increases proportionally;
• The property by which compounds will be created. You can select multiple properties at once:

"QED" (Quantitative Estimate of Drug-likeness) is used to generate molecules that have a high probability of being effective and safe drugs. QED evaluates how well the chemical structure of a compound matches the typical characteristics of successful drugs.

You can set any color for the molecules by clicking on the color square:

A slider is displayed for each property. It is responsible for the degree of qualitative compliance of the created array of molecules with the specified conditions. If the slider is set as far to the right as possible, then the array of compounds will be generated with fewer molecules, but meeting the query as accurately as possible (for example, when creating compounds with low toxicity, the molecules with the lowest toxicity will remain on the map).
The closer the slider is to the left edge, the more options will be generated, but they will match the specified query with less accuracy.
To remove an unnecessary property, click.

After clicking on the “Create” button, the generation of compounds will begin, and gradually new molecules, colored in a different color, will begin to appear on the map. [1]

The generated molecules can also be exported to a separate dataset. To do this, select the required area and export the received data by clicking on the “Save to dataset” button.
The system will prompt you to create a new dataset or add them to an existing one.
A dataset with new molecules will appear in the “Datasets” section.

Hint: If the dataset appears, but there are no molecules in it, click on the button .

Information on working with a dataset (changing the name, selecting, deleting, etc.) is in the “Datasets” section.

In the selected area, click “Create Generator”

7 Section "Reaction prediction"

Experimental section!

The section is devoted to planning the synthesis of chemical compounds and predicting the paths of chemical transformations. In this section you can carry out the following types of predicting:
1.Synthesis — predicting the possible products of a reaction based on the reagents involved in it.
2.Retrosynthesis — predicting reagents to obtain the requested compound. At the moment, a one-stage forecast is available, in the future it is planned to refine this option to a full-fledged retro synthetic analysis.

7.1 Synthesis

To predict possible reaction products based on the reagents involved in it, open the “Planning the synthesis of organic compounds” section, select the “Synthesis” tab and click “+”:

Next, select the molecule input method:

• “SMILES” — when you select this option, a field for entering SMILES appears.
• “Visual editor” — when you select this option, the molecular editor window appears where you can draw a chemical structure.
• “Search in synonyms” — when you select this option, a field for entering synonyms will appear.

After the molecule appears in the preview window, click “Download”.
To model the outcome of a reaction, you need to add at least 2 molecules.

If you want to add an agent, then you should add it here:

Hints: If necessary, you can delete or change the molecule in the graphic editor by moving the cursor to the card and clicking the corresponding button:

• To delete all molecules on the page, click the "Reset" button.

After entering the molecules, click "Prediction".
As we can see, the first result with a prediction reliability score of 0.97 is correct

In positions with the reliability score less than 0.96, unexpected results can be seen, sometimes the model even generates alchemical transformations.
This is not a mistake, but the features of a Transformer-type neural network.

7.2 Retrosynthesis

To obtain the required compound, open the “Planning the synthesis of organic compounds” section, select the “Retrosynthesis”

Next, select the molecule input method:

• “SMILES” — when you select this option, a field for entering SMILES appears.
• “Visual editor” — when you select this option, the molecular editor window appears where you can draw a chemical structure.
• “Search by synonyms” — when you select this option, a field for entering synonyms will appear.
• Syntelly ID – From Syntelly database
• Trivial name – the name of a chemical substance that is historically developed and does not conform to systematic nomenclature
• IUPAC name – The systematic nomenclature developed by the International Union of Pure and Applied Chemistry for naming chemical compounds.[1]

After the molecule appears in the preview window, click “Upload”.

Hints:

• If necessary, you can delete or change the molecule in the graphic editor by moving the cursor to the card and clicking the corresponding button:

• To delete all molecules on the page, click the "Reset" button.

After entering the molecules, click "Predict".

8 Section "Spectra prediction"

This section predicts 3 types of spectra:

1. The Nuclear Magnetic Resonance (NMR) module predicts NMR spectral data (13C, 1H and other nuclei) for small organic molecules. The result is presented as a set of "intensity - chemical shift". Multiplets are also predicted for 1H spectra.
2. The Q ToF MS/MS mass spectrum prediction module for a single ion with an accurate mass. The spectra are calculated for low (10 eV), medium (20 eV) and high (40 eV) energy levels. The result is presented as a set of pairs “exact ion mass – relative intensity”.
3. The IR spectroscopy module allows you to predict the IR spectrum for small organic molecules with various registration options (gas phase, KBr, etc.). The result is displayed as a continuous graph in the axes "wavelength" (in nm) and "relative intensity".

8.1 Nuclear Magnetic Resonance

To predict NMR spectra, you need to open the “Spectra prediction” section and click on the “Nuclear Magnetic Resonance” tab.

1. On the page that opens:

• In the "Original composite structure" field, enter InСhI or SMILES;

–OR–

• Click «Draw molecule» (draw a molecule in the molecular editor).

2. Next, select the NMR method (1H, 13C, 15N, 19F) and click.

The designations 1H, 13C, 15N, 19F refer to specific isotopes of elements that can be predicted:

• 1H (Protium, Hydrogen):
• When this parameter is selected, the module predicts the NMR spectrum based on the most common isotope of hydrogen (1H). This makes it possible to analyze and interpret the chemical environment of hydrogen atoms in a molecule, which is key to understanding organic compounds..
• 13C (Carbon):
• By selecting this option, the user will receive NMR spectrum predictions for the 13C carbon isotope. This helps to investigate the structure of the carbon skeleton of molecules, providing important information about the structure of organic compounds.
• 15N (Nitrogen):
• This selection allows the module to generate NMR spectra predictions for the nitrogen isotope 15N. It is particularly useful for the analysis of nitrogen-containing compounds, including amino acids and proteins, improving understanding of their molecular structure.
• 19F (Fluorine):
• When fluorine 19F is selected, the module provides NMR spectra predictions for that isotope. This is valuable in the study of fluorine-containing compounds, which are widely used in pharmaceuticals and organic chemistry, providing detailed information about the environment of the fluorine atoms.

Hints:

• If you need to change the structure, click the button.
• If you need to reset all configured parameters, click.

Each atom in the molecule is numbered.
The resulting spectrum is displayed in the “Calculated results” field, allowing the user to visually see the characteristic peaks and their distribution.

The "Calculated Results" field displays the spectrum and calculates the "Multiplet" and "Chemical shift" for each numbered atom in the table. Multiplet: For each atom, the multiplet is calculated, which describes the peak splitting pattern in the spectrum. Multiplet indicates the number of neighboring hydrogen atoms (or other nuclei) affecting a given atom, and can be a singlet, doublet, triplet, etc.
Chemical Shift (Meas. Shift): For each atom, a chemical shift is determined that reflects its electronic environment. Chemical shift is measured in frequency units (usually ppm - parts per million) and provides information about the type of chemical bond and electronic structure of the molecule.

The number of rows in the table depends on the number of atoms.

8.2 Mass spectrometry

To predict mass spectra, you need to open the “Spectra prediction” section and click on the “Mass Spectrometry” tab.

1. On the page that opens:

• In the "Original composite structure" field, enter InСhI or SMILES;

–OR–

• Click «Draw molecule» (draw a molecule in the molecular editor).

2. Next, set the parameters to create the spectrum:

• "Spectral type" — a type of mass spectrum. This can be a full spectrum, a fragmentation spectrum, or other specialized types of spectra, each of which provides different information about the molecule.
• "Ion mode" (Positive or Negative) — indicates whether the spectrum will be generated for positively and negatively charged ions. The positive mode is used to search for molecules that easily form cations, and the negative mode is for molecules that form anions.
• "Adduct type" — the type of adduct is important for determining the molecular weight of the target compound. Adducts occur when an ion forms a bond with other substances such as solvents or counterions.
• “Rounding RI” (Retention Index) — the degree of rounding of the retention index in the spectrum. Retention index helps in identifying compounds by comparing their retention time in a chromatograph column with the retention time of known standards.
• "m/z rounding" (mass-to-charge ratio) — the parameter determines how the ratio of the mass of an ion to its charge in the spectrum will be rounded. This is important for accurate identification of molecules and comparison with known standards.

3. After selecting the required options, click the button.

Hints:

• If you need to change the structure, click the button.
• If you need to reset all configured parameters, click.

After the calculations are completed, the following spectra will be displayed on the page:

8.3 Infrared spectrometry

To predict the IR spectrum, open the "Spectra prediction" section and click on the "Infrared Spectrometry" tab.

1. On the page that opens:

• In the "Original composite structure" field, enter InСhI or SMILES;

–OR–

• Click «Draw molecule» (draw a molecule in the molecular editor).

2. Select «Shooting method»:

• "gas (gas phase)" — the method involves measuring the IR spectrum of a substance in the gas phase. The use of a gas phase is preferable to obtain clean and non-overlapping spectral lines. This method is suitable for gaseous samples or for those that can be evaporated without decomposition.
• "liquid (liquid phase)" — in this method, the IR spectrum is taken for a sample in the liquid phase. It is suitable for substances that are liquid at room temperature or for those that can be melted. Liquid samples may require the use of special cuvettes.
• "CCl4" — IR spectroscopy using carbon tetrachloride (CCl4) solvent. Suitable for substances soluble in CCl4, used primarily for analysis of samples that may react or degrade in other solvents.
• "KBr (Potassium Bromide)" — used to create tablets from sample powder mixed with KBr. This method is well suited for the analysis of solid samples that are difficult to evaporate or dissolve.
• "nujolmull (suspension in petroleum jelly oil)" — the sample is mixed with mineral oil to form a paste. This method is suitable for solids that are difficult or impossible to dissolve and that cannot be analyzed as KBr tablets.

3. After selecting the required options, click the button.

Hints:

• If you need to change the structure, click the button.
• If you need to reset all configured parameters, click.

The page will display the calculated spectra and numbering of atoms in the molecule:

9 Section «Synthesis cost»

This section allows you to predict the cost of the final synthetic compounds.
To calculate the data, open the “Synthesis Cost” section.

1. Enter the molecule you want to obtain. To do this, you can enter InChI or SMILES in the “Product” field (or add a molecule using the molecular editor).
2. Optionally add a reagent to the Reagent field by entering InChI or SMILES (or using the molecular editor).
3. In the “Weight in grams” field, enter the desired weight of the synthesized substance in grams.
4. In the “Number of stages” field, enter the number of stages of product synthesis (you can enter up to 6 stages).
5. After selecting the required options, click the button.

Hints:

• If you need to reset all configured parameters, click .
• If you need to change the structure, click the button .
• If the compound contains stereochemistry in SMILES and the algorithm does not find a synthesis path, it is recommended to remove stereochemistry from SMILES. It is possible that synthesis paths suitable for the required substance will be found.
• For industrial synthesis, it is recommended to enter a weight not less than 100g.
• For an initial assessment of the possibility of synthesis, it is recommended to set the maximum number of stages to 6 and the minimum to 1. Perhaps the execution speed will be longer, but the chance of finding reactions is higher. Next, reduce the maximum number of stages or increase the minimum.
• It is not recommended to set the min and max values of the number of stages to the same number, for example min=3 max=3. Perhaps, synthesis paths in 3 stages may not be found, but there are synthesis paths in 2 and 4 stages, which the user can adapt to his specific request.

As a result of the calculation, there are displayed 5 schemes with known reactions to obtain the product entered by the user. To view the received data, select the tab with the diagram.

The page will display the numbered structures of the reactant substances and the substance obtained during the reaction for the selected stage.
The table shows for each substance the name, its estimated cost in dollars, the amount of substance required for the reaction, a possible supplier and a link to confirm the price.
At the bottom of the table the total cost of the product is predicted.

Hints:

• The user can edit any data in the table (including adding a synthesis stage) by clicking on the button.

From the moment you make changes to the table, it is considered edited; it is not saved in your personal account and you need to download it to save the data.

• In the absence of a supplier, the System calculates the approximate cost using artificial intelligence algorithms.
• The results of the selected scheme can be downloaded in .pdf, .csv, .xlsx file formats by clicking the “Download” button at the bottom of the table.

10 Section «PDF to SMILES»

This section contains a tool for automatically recognizing the structural formulas of molecules in documents in .pdf format: patents, scientific articles, test reports, dissertations, etc.

The module provides optical recognition of chemical compound structures and Markush structures.

The module contains 2 tabs, “General” and “Personal”.

The “General” tab contains examples of molecular recognition provided by the Sintelli team; the “Personal” tab contains documents added by the user.

The recognized structures are located on the right side. You can select them and download or save them to a dataset for later work with them..

Hint: To select 1 or more molecules, hold ctrl or shift and left-click on the molecule.

Please note!
You can open the recognized molecule in the molecular editor and make edits to the structures, if there are recognition inaccuracies, by clicking

To upload your document for structure recognition, just drag your file PDF into area
After loading the document, the system starts the process of recognizing structures in it, which (depending on the number of pages and structures) can take from 1 to 8 minutes.
The recognized document looks like this:

11 Section «SMILES to IUPAC»

In this section, you can automatically convert the SMILES string to the IUPAC name.
In the menu on the left, select "SMILES to IUPAC".

In the input field, type the SMILES string and click the "Convert" button.

As a result, the System offers 5 IUPAC name variants.
For each name variant, the prediction accuracy coefficient is indicated:

12 Section «Statistics»

The section contains statistical parameters of the models used in the system.
We adhere to an open policy regarding the metrics of our models. Statistical parameters of models are presented in a separate module on the platform in the “Statistics” section. We use metrics: RMSE, ROC AUC.
RMSE is a metric of root mean square error (the smaller it is, the better). Let's look at the example of paracetamol; the Mouse Intraperitoneal LD50 model predicts a value of 472 mg/kg. If our RMSE metric for the Mouse Intraperitoneal LD50 model is 0.486, this means that the value of the indicator may deviate by half an order of magnitude. How we calculate: we raise 10 to a power (0.486) and multiply/divide the exponent (472) - we get an error from 150 to 1500 mg/kg. This assessment makes it possible to deliberately filter out highly toxic compounds, the values of which are usually below 50 mg/kg.

ROC AUC – if ROC AUC shows 0.5, then this is a random prediction (50/50). The closer the value is to one, the more accurate the prediction. If ROC AUC is equal to 1, this is an ideal model. This indicator is more informative than accuracy, because it takes into account not only the result of the prediction itself, but also the model’s confidence in the prediction.

To view statistics, select the “Statistics” section in the menu on the left.
To display the required parameters, click on the appropriate section.

13 Selecting a language

You can configure the system interface language (Russian or English). To do this, click on the language in the menu on the left and select the required option from the drop-down menu.

14 Bottom profile management menu

In the bottom user management menu, you can change your password or log out, edit your personal data, and also get information about the System you are interested in.

To open this menu, click the button

• “About us” — you follow a link to the site https://syntelly.ru/;
• “We are on Telegram” - you follow a link to telegram channel https://t.me/syntelly, where you can ask a question to the System developers;
• “User Guide”— opens the current instructions for using the System;
• “Profile” — goes to the user’s personal page, where you can change the name, position, organization, as well as change the password for logging into the system or logging out.

That's all for now. We hope this was useful!) If you have any questions, write to us in the tg channel or email admin@syntelly.com.
We promise to help!

next to your username and select the desired item from the drop-down menu: