# Annotate Data and Filter using Annotations #
There are several ways to add annotations to objects in OMERO. Here, addition and filtering of annotations using OMERO.web is described. You can add annotations using the OMERO.web interface to any object(s) that you can select in the left-hand-side tree or central pane, this means Project, Dataset, Image, Screen, Plate and Well.
## Annotate Data manually ##
- **Tags**
Select one or several images from the web client, and in the right-hand pane in {nav General tab}, click the {nav +} button under the {nav Tags} harmonicato add tags manually.
{F18510951}
Select tags from your available tags or create new ones:
{F18510991}
Hit {nav Save}
- **Key-value pairs**
The Key-Value Pairs allow you to add lab-book-like additional metadata for the Image. These Key-Value Pairs are also specifically searchable.
{F18510883}
Select an Image from the web client, and in the right-hand pane in {nav General tab}, click the {nav +} button under the {nav Key-Value Pairs} harmonica to add Key-value pairs manually.
- **Attachments**
Select one or more objects in the left-hand side tree, such as Dataset, Project or Image. Expand the Attachments harmonica in the right-hand pane and click the {nav +} button.
{F18511054}
You can attach any type of file using this function. If you select a file from your local filesystem using the {nav Browse} button, the feature will upload that file to the OMERO.server and save it there. The content of .pdf, CSV and plain text files is also searchable in OMERO.
- ** Others**
You can also add Comments and Rating to selected objects - follow analogous steps to the ones descirbed above for Tags, Key-Value pairs and File Attachments.
## Annotate Data using the auto-tagger##
When an image is uploaded to OMERO, the original filename, including extensions and the path from which it was uploaded, is preserved. It is common for useful metadata to be encoded into the filenames and paths of data. In order to make use of this metadata for searching and organising the data, Auto-tagger enables it to be added as annotations in the form of tags.
###**Terminology**###
• **Tag**: an OMERO text annotation
• **Token**: a section of text which has been automatically parsed from the filename after splitting up the filename by a space, period or underscore. e.g. token1_token2_token3.dv or token3 token4.dv or token5.token6.dv
• **Mapping**: selecting a certain tag to correspond to a certain token, e.g. the tag “Deconvolved” might be created from a token “Deconvolved”.
###**Step-by-step guide**###
1. Select a dataset or image in the data tree. Select {nav Auto Tag} from the view menu in the centre pane toolbar.
{F18511128, size=full}
2. Columns headed with a token name are displayed in the centre pane, with a row representing each image in the dataset.
The initial display shows only existing tags, either already attached to images or not attached but matching a token.
If an image already has a tag applied to it, or if the token for a column is present in this image, checkboxes in the rows are auto-selected. A green background in a cell indicates the image has that tag applied to it. Select the Show All Potential Tags checkbox to generate columns for tokens from the filename path and extensions.
{F18511149, size=full}
3. The drop-down buttons in the headers of the columns are colour-coded as:
- Pink (empty drop-down button) - indicating a token which did not match any existing tags in the system, or there were multiple matching tags and there was no way to make a determination as to which was intended.
- Blue (tag name in drop-down) - indicating a token which matched exactly one existing tag and has been automatically mapped to this tag. Blue columns are also shown when the user manually applies a mapping.
- Green (tag name in drop-down) - indicating a tag which is already applied to some images in this dataset, but does not match any of the tokens.
{F18511188, size=full}
4. The default rarity threshold setting is 2. Use the Rarity Threshold slider to adjust the threshold to a higher value to refine the display of tokens according to how commonly they were detected on images. Click the Apply button to save the changes. Click the Refresh button to reset the table to its original state without applying changes.
(IMPORTANT) Tokens that are only numbers are ignored, as these are usually very numerous and it is impossible to intelligently map them at this time.
The filename is tokenised using a **space**, **period** or **underscore** as the splitter. **Dashs** are not recognized as splitters.
{F18511225, size=full}
5. Click on the Pink button drop-down to add a suggested tag. Select {nav New/Existing Tag}. Edit or accept suggested tag name. Add a description if required.
Click {nav OK}. Once happy with the suggestions indicated by the boxes ticked in the centre pane, click the {nav Apply} button, which will attach the tags to the images.
{F18511247, size=full}
## Filter using annotations ##
Images can be filtered by Name, Tag, Key-Value pairs or Rating in the centre pane, using the {nav Add filter} chooser above the thumbnails.
{F18511262, size=full}
## Remove an annotation ##
Find the annotation you want to remove. Click on the minus sign to the right of it.
The removal action just unlinks the File Attachment from the selected object(s). The File Attachment is not deleted from the server. If deletion is needed, click in the workflow above on the **cross icon** instead of the **minus icon**.