<h1 id="2024-01-15">2024-01-15<a aria-hidden="true" class="anchor-heading icon-link" href="#2024-01-15"></a></h1>
<h1 id="this-is-marcos-daily-open-notebook">This is Marco's daily open-notebook.<a aria-hidden="true" class="anchor-heading icon-link" href="#this-is-marcos-daily-open-notebook"></a></h1>
<p>Today is 2024.01.15</p>
<h2 id="notes">Notes<a aria-hidden="true" class="anchor-heading icon-link" href="#notes"></a></h2>
<h3 id="todays-call-with-luca">Today's call with Luca<a aria-hidden="true" class="anchor-heading icon-link" href="#todays-call-with-luca"></a></h3>
<p>We don't have negatives. --> could use the weight of the loss for the negatives </p>
<p>even having 2 models is not a big problem. we could have a convolution that parses those features</p>
<p>we need : </p>
<ul>
<li>features for molecules : </li>
<li>features for species</li>
</ul>
<p>It looks like CHEM-BERT works better than MAP4 for the features.</p>
<h2 id="todo-today">Todo today<a aria-hidden="true" class="anchor-heading icon-link" href="#todo-today"></a></h2>
<p>Read papers : </p>
<ul class="contains-task-list">
<li class="task-list-item"><input type="checkbox" checked disabled> <a href="https://static.googleusercontent.com/media/research.google.com/it//pubs/archive/40671.pdf">https://static.googleusercontent.com/media/research.google.com/it//pubs/archive/40671.pdf</a></li>
<li class="task-list-item"><input type="checkbox" disabled> <a href="https://vigna.di.unimi.it/ftp/papers/HyperBall.pdf">https://vigna.di.unimi.it/ftp/papers/HyperBall.pdf</a> </li>
</ul>
<h2 id="doing">Doing<a aria-hidden="true" class="anchor-heading icon-link" href="#doing"></a></h2>
<h2 id="done">Done<a aria-hidden="true" class="anchor-heading icon-link" href="#done"></a></h2>
<ul>
<li></li>
</ul>
<h2 id="todo-tomorrow">Todo tomorrow<a aria-hidden="true" class="anchor-heading icon-link" href="#todo-tomorrow"></a></h2>
<ul class="contains-task-list">
<li class="task-list-item"><input type="checkbox" checked disabled> insert nodes for each term in NCBI taxon. </li>
<li class="task-list-item"><input type="checkbox" checked disabled> Insert NCBi taxon to the graph that I created. This will allow to increase the information for the species with an NCBi id and the other will just have the wikidata topology.</li>
</ul>

2024-01-15


# Overview

The Digital Botanical Gardens Initiative (DBGI, www.dbgi.org) ambitions to explore innovative solutions for the collection, management and sharing of digital information acquired on living botanical collections. A particular focus will be placed on the large scale characterization of the chemodiversity of living plants collections through mass spectrometric approaches. The acquired data will be structured, organized and connected with relevant metadata through [semantic web](https://en.wikipedia.org/wiki/Semantic_Web) technology. The gathered knowledge will then inform ecosystem functioning research and orient [biodiversity conservation](https://en.wikipedia.org/wiki/Conservation_biology) projects. The DBGI initially aims to take advantage of the readily available living collections of Swiss botanical gardens to establish robust and scalable biodiversity digitisation workflows. The ultimate goal is to apply these approaches in the field and at the global scale in wildlife ecosystems. 

# Main goals

The DBGI main goals are briefly summarized below :

![](/assets/images/2022-06-18-17-13-23.png){max-width: 400px, display: block, margin: 0 auto}

# An Open Science project


## Working with the workshop doors open

![](/assets/images/2022-06-18-17-47-56.png){max-width: 400px}

A way of working briefly described by Andy Matuschak here https://notes.andymatuschak.org/z21cgR9K3UcQ5a7yPsj2RUim3oM2TzdBByZu. 


## Open Notebook Science

If we apply this "working with the workshop doors open" idea to a scientific research endeavor we understand it as respecting the core concepts of Open Science. These is more precisely the principle of Open Notebook Science https://en.wikipedia.org/wiki/Open-notebook_science

For the projects of the DBGI we will follow such an Open Notebook Science approach in order to share early ideas and results but also the less shiny parts of the research projects. 

You can follow the DBGI related Open Notebooks here [[open-notebook]].

Expect to find strange ideas, badly formulated text and poorly written code. However if you have time and interest you should be able to participate at each step of the process. Hopefully this will lead to enhanced collaboration and participation. Maybe yes, maybe not. Let's make it a methodological experiment !

Each pages of this website are written in markdown and hosted on Github. You can edit them by clicking on the "Click here to edit this page on Github !" link at the bottom of the page. See down there 👇

If you are willing to participate the the DBGI Dendron notes have a look at the [[welcome]] section to get you started !


## What's new ?!

This web site might not necessarily seem to be change since the last time you came to visit.

If something moved recently you should be able to track this over there https://github.com/digital-botanical-gardens-initiative/dendron-dbgi/commits/main