Spectroscape User Manual with GIFs

Search query

Each MS2 spectrum in the archive is designated with an identifier (id). The identifier ranges from 0 to size of the archive (exclusive). The following example shows how to search an query spectra from NIST library. The query id is 25757878.

A generic square placeholder image with rounded corners in a figure.
Fig.1 - library entry as a bridge of two clusters.

Tooltip on node and edges, PSM visualization and search new query

Mouse over any node or edge to show the tooltip information. Single click on any node to show the PSM on the right side using Lorikeet. Double click on any node to start a new search using that node as query spectrum.

A generic square placeholder image with rounded corners in a figure.
Fig.2 - Tooltip on node and edges. Single click to show the PSM. Double click to start a new search.

Tuning parameters of graph

Change the strength of repulsive force between nodes and the attractive force (gravity) to the center. Also adapt the width of edges.

A generic square placeholder image with rounded corners in a figure.
Fig.3 - Tuning the parameters of the graph, repusive and attractive force, and edge width.

Investigate unidentified PSM corresponding to semitryptic peptide

The following examples shows how to find the correct solution of an unidentified peptide. The query id is 18497303 . In this case, the query node is in black color which means it is not identified. However, there are many neighbors in red are identified as DDPLTNLNTAFDVAEK, and the highlighted edge suggests that the delta mass is 230Da, which equals to mass of "DD". So we can guess if we remove the DD from "DDPLTNLNTAFDVAEK", the remaining part "PLTNLNTAFDVAEK" will match to the query node. In the PSM viewer, we edit the peptide and click update button, and see it is a very good PSM. The user can send a message to the server to report this error.

A generic square placeholder image with rounded corners in a figure.
Fig.4 - Investigate into the unidentified PSM using the information of its neighbors. The highlighted edges shows the mass difference of 230 Da equals to the mass of "DD". And we find the red nodes are all matched with a peptide that starts with "DD". What if we remove the DD and use the remaining part to match with the query spectrum? Yes, it turns out that the correct anwser is "PLTNLNTAFDVAEK".

Investigate into mixture spectra of two similar peptides

The following example shows two similar peptides co-fragmented into one MS2 spectrum. The query id is 18793053 . In this case, the neighbors of the query node form one tightly connected cluster. All of them are identified, but as two different peptides, ADIDVSGPK and ADVDISGPK. Thse two peptides share many fragment ions. From the back-to-back PSM plot implemented in Spectroscape interface, one can clearly see the query spectra contains peaks of both peptides.

There are four different mode when it comes to the back-to-back plot: query, neighbor, separate and mixture. In the first three mode, the back-to-back plot shows two spectra, the top spectrum is the one user selected by single clicking on the neighbouring node, and the bottom one is the query spectrum. The query method will annotate the two spectrum with the peptide of query node. The neighbor method will annotate the two spectrum with the peptide of the user selected spectra. The separate method will annotate the two spectra with their own peptide. The last method, the mixture annotation will show top and bottom with the same query spectrum, but annotated with query peptide (bottom) and peptide of user selected node (top).

A generic square placeholder image with rounded corners in a figure.
Fig.5 - Investigate into a mixture spectra. The cluster returned by Spectroscape contains two different peptides. According to the analysis with back-to-back PSM plot, we can tell the query node is a mixture of peptide ADIDVSGPK and ADVDISGPK .

Search a peak list out of archive

It is possible to search a peak list out of the archive, the only information we need is a peak list. The peak list should contain multiple rows, each row is a pair of values, mz and intensity. Here is an example peak list.

120.058 48
156.100 35
175.113 39
201.114 50
207.095 36
214.054 38
229.099 43
258.122 32
298.131 31
303.105 30
312.169 25
333.135 23
362.188 46
379.156 26
425.238 21
433.234 45
438.178 22
441.169 28
466.163 29
496.284 27
566.293 18
583.322 47
596.292 42
682.383 41
759.350 37
765.423 16
783.429 49
873.396 33
893.477 19
894.453 34
911.482 44
974.441 24
1044.450 17
1058.549 20
1061.479 40
1145.586 15

This Peak list above actually is the query spectrum 21958751. Now we can open the Spectroscape web interface, choose to search a peak list, then we can paste the peaks above to the text box and click button "search peak list". It's done. The web interface will show the network between the query spectra and its neighbors. Note that this spectrum is out of the archive, therefore its query id become "-1".

A generic square placeholder image with rounded corners in a figure.
Fig.6 - Search a peak list. It turns out that this is a mixture spectrum. It is actually from the archive, where the query id is 21958751.

©2022-2023 Spectroscape | Design: Long WU @ HKUST in Henry Lam Group | Source code on GitHub