The GNPS molecule explorer builds upon the continuous identifications of all the public GNPS datasets in MassIVE. Now is a central location where one is able to look at all this data from a compound centric point of view instead of a dataset centric point of view. This allows users to ask the questions, "where have I seen this molecule before?" and "are there any analogs to this compound seen in the data?". The molecule explorer view can be found as a link from this particular page on GNPS here.
The molecule explorer interface has several columns complete with descriptions below.
|Compound||The common name of the compound.|
|NeighborCount||The total number of spectra that are putative analogs of the compound across all datasets|
|UnidentifiedNeighborCount||The total number of spectra that are putative analogs of the compound across all datasets that have not been identified.|
|PrecursorNeighbor||The total number of unique precursor neighbors per dataset, summed across all datasets|
|UnidentifiedPrecursorNeighbor|| The total number of unique precursor neighbors per dataset where all spectra with all membership spectra not having a single identification, summed across all datasets
|View Matches and Neighbors||See Below|
This particular column has several links in each cell that merits further explanation. In this cell each row represents the identifications and putative analogs to a particular dataset. The very first link in the row is a link to the identifications of the particular compound from the dataset. The link name is the ID number of the dataset, and the alt text is the name of the dataset. The second link is a link to the putative analogs that are not identified. The name of the link is the number of unique unidentified precursors from the particular dataset.
The molecule explorer enables the exploring the molecular space using mass spectrometry. One of the main problems in natural product datasets is the low rates of identification of the spectra. To help to solve this problem, the GNPS spectral libraries need to grow. Even for commonly seen compounds many of the analogs of those compounds are not present in the spectral libraries. Since the molecule explorer allows users to look at the unidentified putative analogs and so will aid in the expansion of coverage of molecular families.
By clicking the "View X Neighbors" link, shown here:
users will be taken to a page showing all unidentified putative analogs of the originating molecule for a particular dataset. In this example, clicking the "View 4 Neighbors" in the first row will take the user to the page showing the putative unidentifed analogs for Surfactin the MassIVE dataset MSV000078602 (GNPS - Streptomyces Extracts). It will take users to this page as an example here .
To further explore these putative identifications, one can click the double down arrow shown here:
to show the neighbors of this particular node. One of these neighbors should be identified with the molecule that user was originally exploring. In this example below:
it is shown that this particular putative analog of Surfactin is indeed both a neighbor of Surfactin and Surfactin C14. It also possible to plot the two related spectra to compare them visually by clicking the spectrum icon on the left of the row. The image on the left is of the putative analog, and the image on the right is the identified Surfactin spectrum:
If the user is confident that a particular putative analog is indeed an analog, the user is able to add the particular spectrum to the GNPS libraries. By clicking the AddToLibrary link a user is prompted to select a specific experimental spectrum instead of a consensus spectrum to add to the library. To proceed from this point, please refer to this page .