Please note: since the last update we have discovered some bugs and inaccuracies in the function of SINimilarity. We are working hard to solve the problems. If you have questions, or get strange results from your search, please contact us at firstname.lastname@example.org.
How to use SINimilarity
The SINimilarity tool has a background database of more than 500 000 identified chemical structures, whereof approximately 90 000 have CAS or EC numbers available. This means that you can search for any of those structures, names or CAS/EC numbers and see if the substance you are looking for is similar to a substance on the SIN List.
For the advanced user there is also the possibility to enter any structure of a chemical as a SMILES notation. This could be useful, e.g. if you are investigating a substance without a CAS/EC number or with an ambiguous name. The SMILES search function is dynamic so most, if not all, versions of SMILES are accepted.
SINimilarity looks for two things in the molecular structure:
1. The overall structural resemblance to SIN List substances. You will see which, if any, substances on the SIN List that has 70% or more structural resemblance to your substance. Matches with the SIN List show in orange colour. You can also click to view the corresponding substances in the SIN database.
2. Specific elements of importance to the toxicity. These elements are the same as the ones used to group the SIN List. If your substance contains this structural element, you will see the name of the corresponding SIN group.
If a substance turns up in yellow, it does not reach 70% resemblance, but may still contain the structural element, which is then indicated by displaying the group name.
How to interpret the result?
SINimilarity provides a first good indication and can be helpful as a first screening tool. An answer from SINimilarity, indicating similar structure and/or grouping, is a very good reason to ask for more information; given the structural similarity, have the corresponding hazardous properties been carefully investigated?
If a substance both has the similar structure and the structural element there is a good reason to proceed with additional caution. If you have only either resemblance or the structural element, the result is more ambiguous to interpret. A substance with similar structure, but without the specific structural element could, after careful investigation, turn out to be a functional alternative to a known hazardous chemical.
Please note that SINimilarity builds on structural similarity only. To verify potential hazardous properties, further evaluation is needed. If a chemical has no similarity to SIN chemicals it can still have hazardous properties.
SINimilarity is based on the division of the SIN List into 31 groups. The base for the grouping is specific molecular structural elements, which have been identified to be important for the toxicity of the substance. Some chemicals on the SIN List belong to more than one group, as they contain more than one of these structural elements. About 60 substances on the SIN List do not belong to a group. For more detailed information on the methodology for the grouping and for SINimilarity, please consult the methodology document.
The SIN List is divided into the following groups:
- Amino carbonyl compounds
- Antimony compounds
- Aromatic amines
- Arsenic compounds
- Azo compounds
- Beryllium compounds
- Boron compounds
- Cadmium compounds
- Chromium compounds
- Cobalt compounds
- Glycol ethers
- Lead compounds
- Mercury compounds
- Mineral fibres
- Nickel compounds
- Nitro compounds
- Organotin compounds
- Perfluorinated compounds
- Polyhalogenated aromatics
- Polyhalogenated alkanes
- Polyhalogenated alkenes
- Thioamino carbonyl compounds
The SIN groups “Petroleum” and “Mineral fibres” contain substances of very complex chemical composition. Substances in these groups are for this reason not used in the SINimilarity tool. Inorganic compounds and many salts are not suited for the similarity methods used in SINimilarity. This is especially true for many compounds in the metal groups. The similarity will be too low to be shown. If the substance contains a group specific metal, it will be identified with regard to that group, which is most often the most useful information on metals.