Home

Welcome to the Wiki of the amazing Nala pipeline.

You can have a quick peek on why Nala was created in the first place. --> 2 theses' project in order to build up tagtog.net

Introductory talk:
Online presentation Thesis

View the pipeline visualization externally in a bigger resolution here. Editable version of the online diagram can be found here. (requires to log in/create an account to Lucidchart).

Goals of 2 theses and this method:

1. Study significance of NL mentions in mutation mention recognition

• ratio of standard vs NL in abstracts & full text
• % of novel mutations not present in SwissProt (would require manual annotation of protein
• % of mutation mentions in natural language that don't appear as standard mention

2. Define/extend corpus of NLs

• size depends on significance of NLs

3. Method for mutation mention extraction grounded to their genes/proteins

• Mutation mention recognizer better than tmVar for standard mentions
• If NLs are relevant, prove good F1 performance (> 70-80)
• Simple or optionally advanced normalization method
• Easy to use program:
  • Good documentation:
    • code
    • end-user (biology researcher level, how to call from the command line, ...)
  • Accept inputs: programmatical call (string), text file, corpora' formats**
  • Accept outputs: ann.json (tagtog suitable)

4. Paper

• Full draft (1 or 2 papers?) by end of August submittable to Burkhard Rost
• Submit by September-October

Theses Documentation

• Thesis Carsten
• Thesis Aleksander