Moving beyond MARCO.
The use of imaging systems in protein crystallisation means that the experimental setups no longer require manual inspection to determine the outcome of the trials. However, it leads to the problem of how best to find images which contain useful information about the crystallisation experiments. The...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Public Library of Science (PLoS)
2023-01-01
|
| Series: | PLoS ONE |
| Online Access: | https://doi.org/10.1371/journal.pone.0283124 |
| _version_ | 1797843071695585280 |
|---|---|
| author | Nicholas Rosa Christopher J Watkins Janet Newman |
| author_facet | Nicholas Rosa Christopher J Watkins Janet Newman |
| author_sort | Nicholas Rosa |
| collection | DOAJ |
| description | The use of imaging systems in protein crystallisation means that the experimental setups no longer require manual inspection to determine the outcome of the trials. However, it leads to the problem of how best to find images which contain useful information about the crystallisation experiments. The adoption of a deeplearning approach in 2018 enabled a four-class machine classification system of the images to exceed human accuracy for the first time. Underpinning this was the creation of a labelled training set which came from a consortium of several different laboratories. The MARCO classification model does not have the same accuracy on local data as it does on images from the original test set; this can be somewhat mitigated by retraining the ML model and including local images. We have characterized the image data used in the original MARCO model, and performed extensive experiments to identify training settings most likely to enhance the local performance of a MARCO-dataset based ML classification model. |
| first_indexed | 2024-04-09T16:58:33Z |
| format | Article |
| id | doaj.art-cc833f1c5b834feea6d1e9fb49c3755b |
| institution | Directory Open Access Journal |
| issn | 1932-6203 |
| language | English |
| last_indexed | 2024-04-09T16:58:33Z |
| publishDate | 2023-01-01 |
| publisher | Public Library of Science (PLoS) |
| record_format | Article |
| series | PLoS ONE |
| spelling | doaj.art-cc833f1c5b834feea6d1e9fb49c3755b2023-04-21T05:34:30ZengPublic Library of Science (PLoS)PLoS ONE1932-62032023-01-01183e028312410.1371/journal.pone.0283124Moving beyond MARCO.Nicholas RosaChristopher J WatkinsJanet NewmanThe use of imaging systems in protein crystallisation means that the experimental setups no longer require manual inspection to determine the outcome of the trials. However, it leads to the problem of how best to find images which contain useful information about the crystallisation experiments. The adoption of a deeplearning approach in 2018 enabled a four-class machine classification system of the images to exceed human accuracy for the first time. Underpinning this was the creation of a labelled training set which came from a consortium of several different laboratories. The MARCO classification model does not have the same accuracy on local data as it does on images from the original test set; this can be somewhat mitigated by retraining the ML model and including local images. We have characterized the image data used in the original MARCO model, and performed extensive experiments to identify training settings most likely to enhance the local performance of a MARCO-dataset based ML classification model.https://doi.org/10.1371/journal.pone.0283124 |
| spellingShingle | Nicholas Rosa Christopher J Watkins Janet Newman Moving beyond MARCO. PLoS ONE |
| title | Moving beyond MARCO. |
| title_full | Moving beyond MARCO. |
| title_fullStr | Moving beyond MARCO. |
| title_full_unstemmed | Moving beyond MARCO. |
| title_short | Moving beyond MARCO. |
| title_sort | moving beyond marco |
| url | https://doi.org/10.1371/journal.pone.0283124 |
| work_keys_str_mv | AT nicholasrosa movingbeyondmarco AT christopherjwatkins movingbeyondmarco AT janetnewman movingbeyondmarco |