From b62cbe86926fc41be8d1d5816681d101dfbc9a98 Mon Sep 17 00:00:00 2001 From: rautenberg Date: Thu, 29 Sep 2016 09:39:11 +0200 Subject: add athens workshop 2016 to presentations --- .../enanomapper-athens16/athens-workshop.html | 232 +++++++++++++++++++++ views/presentations.haml | 5 +- 2 files changed, 236 insertions(+), 1 deletion(-) create mode 100644 public/presentations/enanomapper-athens16/athens-workshop.html diff --git a/public/presentations/enanomapper-athens16/athens-workshop.html b/public/presentations/enanomapper-athens16/athens-workshop.html new file mode 100644 index 0000000..a2afd00 --- /dev/null +++ b/public/presentations/enanomapper-athens16/athens-workshop.html @@ -0,0 +1,232 @@ + + + + + + + + + Filling knowledge gaps in Nanotoxicology with Read Across Predictions, Practice & Requirements + + + + + + +
+

Filling knowledge gaps in Nanotoxicology with Read Across Predictions, Practice & Requirements

+

+Christoph Helma, Micha Rautenberg, Denis Gebele +

+

in silico toxicology gmbh, Basel, Switzerland
+
+
+

+
+
+

Contents

+ +
+
+

lazar read across framework

+

A reproducible version of the read across procedure commonly used in toxicological risk assessment (based on the k-nearest-neighbor algorithm)

+ +

lazar was originally designed for small molecules with a defined chemical structure. The nanoparticle extension was developed and validated within the eNanoMapper project.

+
+
+

Similarity calculation

+
+
Requirements
+
Descriptors (features) for the query substance and the neighbor candidate +
+
Observation
+
A large number of irrelevant features can lead do meaningless similarity estimates +
+
Relevant features
+
Features that correlate significantly with toxicity (Pearson correlation p-value < 0.05) +
+
Weighted cosine similarity
+
    +
  • Scaled and centered relevant feature vectors
    • +
  • Feature contributions weighted by Pearson correlation coefficient
    • +
  • Similarity threshold: sim > 0.5
    • +
+
+
+
+
+

Local regression algorithms

+ +

Partial least squares and random forest models use the caret R package with default settings

+

Prediction intervals: 1.96*RMSE of carets bootstrapped model predictions

+

If PLS/RF modelling or prediction fails, lazar resorts to using the weighted average method.

+
+
+

Validation

+ +
+
+

Data requirements

+ +

Net cell association endpoint of the Protein corona dataset (121 gold and silver particles)

+
+
+

10-fold crossvalidations

+ +++++++ + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
DescriptorsAlgorithmr2RMSE
PhyschemWA0.42, 0.46, 0.482.02, 1.94, 1.92
PhyschemPLS0.53, 0.54, 0.491.83, 1.8, 1.9
PhyschemRF0.53, 0.52, 0.541.82, 1.84, 1.79
ProteomicsWA0.66, 0.63, 0.63 *1.58, 1.62, 1.66 *
ProteomicsPLS0.59, 0.66, 0.63 *1.74, 1.56, 1.65 *
ProteomicsRF0.66, 0.65, 0.63 *1.56, 1.59, 1.64 *
AllWA0.73, 0.66, 0.66 *1.41, 1.57, 1.58 *
AllPLS0.67, 0.64, 0.69 *1.53, 1.63, 1.5 *
AllRF0.69, 0.69, 0.7 **1.51, 1.5, 1.46 **
+

Gold and silver particles included!

+
+
+

Correlation plot

+
+Correlation of log2 transformed net cell association measurements with random forest predictions using physchem properties and protein corona data. +

Correlation of log2 transformed net cell association measurements with random forest predictions using physchem properties and protein corona data.

+
+
+ +
+

Exercises

+

Try the nano-lazar versions at

+
+
Old (stable) version (physchem only)
+
https://nano-lazar.in-silico.ch +
+
Next release
+
https://nano-lazar-dev.in-silico.ch/predict +
+
+
+
+

Questions

+ +
+ + diff --git a/views/presentations.haml b/views/presentations.haml index 93384dc..6e39eb7 100644 --- a/views/presentations.haml +++ b/views/presentations.haml @@ -1,6 +1,9 @@ View recent presentations and slides : %ul - %li + %li + %a{:href => "/presentations/enanomapper-athens16/athens-workshop.html", :rel => "external"} Filling knowledge gaps in Nanotoxicology with Read Across Predictions, Practice & Requirements + (eNanoMapper Athens 2016) + %li %a{:href => "/presentations/enanomapper-nano-lazar/enm-presentation.html", :rel => "external"} Validation of read across predictions for nanoparticle toxicities (eNanoMapper 2016) %li -- cgit v1.2.3