diff --git a/images/0_bot.png b/images/0_bot.png
new file mode 100644
index 0000000000000000000000000000000000000000..6e60998d9f9a9a65d31bb6c3986fe58adeb51d82
Binary files /dev/null and b/images/0_bot.png differ
diff --git a/images/2_2_move.png b/images/2_2_move.png
new file mode 100644
index 0000000000000000000000000000000000000000..d49771d7429e7773d790f9826989eeb5a38dfd36
Binary files /dev/null and b/images/2_2_move.png differ
diff --git a/images/2_3_complete.png b/images/2_3_complete.png
new file mode 100644
index 0000000000000000000000000000000000000000..d9b7fc448b3b032eaefab771cd29e3bc61bb167a
Binary files /dev/null and b/images/2_3_complete.png differ
diff --git a/images/2_4_reaction.png b/images/2_4_reaction.png
new file mode 100644
index 0000000000000000000000000000000000000000..8a08e82a28c83fd7083c040a270165fe928beb12
Binary files /dev/null and b/images/2_4_reaction.png differ
diff --git a/images/2_5_fact_p1.png b/images/2_5_fact_p1.png
new file mode 100644
index 0000000000000000000000000000000000000000..be74c538a9d4f069bfdea4a32312025f13dbba51
Binary files /dev/null and b/images/2_5_fact_p1.png differ
diff --git a/tm_curation/curate_from_tm.md b/tm_curation/curate_from_tm.md
index c31b95968fa37332aa027f7e16d9bab8f108a270..edf7975da06a8916a1d4e6818649abad23e3ca05 100644
--- a/tm_curation/curate_from_tm.md
+++ b/tm_curation/curate_from_tm.md
@@ -5,10 +5,24 @@ This excercise will cover:
- building a simple fact using selected entries
- building an advanced fact
+## Read first: vocabulary and data model
+
+BioKC relies on [SBML (Systems Biology Markup Language)](https://sbml.org) as its underlying data model and follows [annotation qualifiers](http://mbine.org/standards/qualifiers) for its annotations. This influences the vocabulary of BioKC.
+
+Important vocabulary disambiguations:
+- **compartment** a physical or conceptual location of elements of interactions
+- **species** are elements of interactions; in SBML every species has an assigned compartment
+- **reactions** describe interactions between interacting elements: **reactants** and **products**
+- **modifiers** are elements of a reaction (interaction) with a specific role; in SBML a reaction can have an element that modifies them.
+
+In SBML, a reaction (interaction) must have at least one reactant (input) and at least one product (output).
+
+In SBML, a reaction can have multiple reactants, products and modifiers. Reaction type is the same for reactants and products. Each modifier can have a different role (type of modifying relationship).
+
## Table of contents
- [Literature collection](#literature-collection)
- [Step 1: Import sentences](#step-1-import-sentences-from-the-data-set)
-- [Step 1: Curate from basket](#step-2-build-facts-from-entries-in-the-basket)
+- [Step 2: Curate from basket](#step-2-build-facts-from-entries-in-the-basket)
## Literature collection
@@ -38,41 +52,76 @@ The file contains a table with the following columns:
### 1.2. Import sentences from a file
-1.2.1 Choose the "indra_biokc.tsv" file
+1.2.1 Choose the "indra_biokc.tsv" file (1 in the figure below)
-1.2.2 Explore sentences from the file are imported and shown in the area below
+1.2.2 Explore sentences from the file are imported and shown in the area below (2 in the figure below)
### 1.3. Filter by keywords
-1.3.1 Type the keyword you want to filter entries by; this can be a publication identifier, molecule identifier or a term in the sentence.
+1.3.1 Type the keyword you want to filter entries by; this can be a publication identifier, molecule identifier or a term in the sentence. Example: use "tnf" as a keyword (1 in the figure below).
Important to note:
- text based searches are less precise ("tnf" query will have more results than "P01375")
- multiple keywords can be combined wtih space as a separator (e.g. "TNF IKK")
-1.3.2 Notice that the number of results is dynamically reduced
+1.3.2 Notice that the number of results is dynamically reduced (2 in the figure below)
### 1.4. Add entries to the "Curator's basket"
-1.4.1 Select entries you want to curate; metadata is fetched for publication identifiers
+1.4.1 Select entries you want to curate; metadata is fetched for publication identifiers (1 in the figure below)
-1.4.2 Notice that data from selected entries is added to the "Curator's basket"
+1.4.2 Notice that data from selected entries is added to the "Curator's basket" (2 in the figure below)
-1.4.3 Click "Go to basket" to start curation using the selected entries
+1.4.3 Click "Go to basket" to start curation using the selected entries (3 in the figure below)
## Step 2: Build facts from entries in the Basket
+Take a look at the [vocabulary disclaimer](#read-first-vocabulary-and-data-model) to read about species (elements), compartments (location) and reactions (interactions).
+
### 2.1. Initial steps
-2.1.1 In the "Curator's basket", notice the entries selected in the previous step
+2.1.1 In the "Curator's basket", notice the entries selected in the previous step (1 in the figure below)
+
+2.1.2 Set the fact group to "Training"; this is where your facts will be saved (persisted) later (2 in the figure below)
+
+2.1.3 Set up the initial compartment for the facts, use  button to add a compartment and  to edit it. In the example below, the compartment is set to "cell" with the [qualifier annotation](http://mbine.org/standards/qualifiers) set to "bqbiol:is", annotation ontology set to "Cell Ontology" and the assigned identifier to [CL:0000000 (cell)](http://purl.obolibrary.org/obo/CL_0000000) (3 in the figure below).
+
+
+
+### 2.2. Add elements
+
+2.2.1 Drag and drop elements of an interaction. Example: drag "TNF" and "p65".
+
+
+
+2.2.2 Cell type and element type need to be defined for the imported elements.
+
+Example:
+- use previolusly defined "cell" as the "Compartment"
+- set "Bioentity type" to "Generic protein"
+- set identifier namespace to "Uniprot Knowledgebase".
+
+Repeat for p65 and set its name to "RELA" by modifying the field "Species Name".
+
+
+
+2.2.3 With defined elements, create an interaction: use  button to add a compartment and  to edit it. Complete the information to create a simple interaction representing positive influence of TNF on RELA (p65).
+
+- use previolusly defined "cell" as the "Compartment"
+- set "Reaction type" to "Positive influence"
+- set "Reactant" to "TNF" and "Product" to "RELA".
+
+
+
+2.2.4 With defined interaction, create a fact and annotate it with the sentence. Use  button to add a fact. Then:
-2.1.2 Set the fact group to "Training"; this is where your facts will be saved (persisted) later
+1. Drag created reaction to the reaction area of the fact.
-2.1.3 Set up the initial compartment for the facts, use  button to add a compartment and  to edit it. In the example below, the compartment is set to "cell" with the [qualifier annotation](http://mbine.org/standards/qualifiers) set to "bqbiol:is", annotation ontology set to "Cell Ontology" and the assigned identifier to [CL:0000000 (cell)](http://purl.obolibrary.org/obo/CL_0000000).
+2. Drag the corresponding field to the evidence area of the fact.
-
\ No newline at end of file
+