Greetings, please give a patient hearing to the below message -
Date of acceptance - 14 November 2024
Journal - Integrative Biology (Oxford University Press)
Title - Machine learning ranking of plausible (un)explored synergistic gene combinations using sensitivity indices of time series measurements of Wnt signaling pathway
Link - https://doi.org/10.1093/intbio/zyae020
Background -
The published research develops a search engine that reveals ground breaking results in the form of higher order (un)explored/(un)tested combinations of genes/proteins (as biological hypotheses), based on sensitivity indices that capture the strength of influence of factors (here genes/proteins) that affect the Wnt signaling pathway. The pipeline uses kernel based sensitivity indices to capture the influence of the factors in a pathway and employs powerful support vector ranking algorithm. Because of the above point, biologists/oncologists will be able to narrow down their search to particular combinations that are ranked and if a synergistic functioning is confirmed, will be able to study the mechanism between the components of a combination, in the Wnt pathway. The search engine design is not only limited to one dataset and a range of combinations of genes/proteins. The framework can be applied/modified to all problems where one is interested in searching for particular combinations of factors involved in a particular phenomena. The published work combines work presented at EMBO-Wnt 2016 conference (Brno, Czech Republic) and Berkeley Cell Symposia : Technology. Biology. Data Science. 2016 (Berkeley, USA)
In my limited opinion the following goals can be achieved using the published work (See demonstration after this section) -
Biologist/oncologist can download the available code in R (and with support from any personnel having R programming experience), and apply to their data sets to rank/prioritise unknown combinations of genes/proteins which might be working synergistically in a pathway.
Because of 1. biologist/oncologist will not have to struggle to search for combinations of interest. The rankings of the combinations shed light on how the combinations can be searched/located. Probably, this work will make life easier.
Based on these rankings, the modifications of the work will help in writing grant proposals for testing machine learning based discoveries. This will also help biologist/oncologist get financial support for their research.
Combinations of any order can be generated and ranked. However, computational resources might be required and the search engine might have to be fine tuned.
Finally, the work might help answer many questions in cell biology. Though experimental tests need to be conducted on the discoveries made by applying the above machine learning based pipeline.
Code available for download -
Link to download the R code has been made available in the above publication. Please cite the above published manuscript (of Oxford University Press) for using this code, if using for research publications.
To demonstrate what can be done, here is an example for a static data -
The ETC-1922159 was released in Singapore in July 2015 under the flagship of the Agency for Science, Technology and Research (A*STAR) and Duke-National University of Singapore Graduate Medical School (Duke-NUS). In the publication in https://www.nature.com/articles/onc2015280 , recording of regulation (up/down) of some 5000 genes were made (available online with the published paper), after the drug was tested on Colorectal cancer cells. I tested the modification of the search engine (as mentioned in the above publication in Oxford University Press) and discovered various 2nd order combinations of genes that might affect various pathways, after the drug was administered. The documentation of the discoveries cite published confirmatory results of existing gene combinations. Next, based on these confirmatory results, we find appropriate rankings that the search engine points to for the ETC-1922159 data. Lastly, based on these rankings, we infer new combinations pointed out by the search engine. Some of the results of the work were presented in a poster, in the first Gordon Wnt Research Conference in 2017, in Vermont in USA. The discoveries have been segregated into 8 areas for research. These are documented in 8 different articles, unpublished preprints of which have been made available in the following links -
- Wnt related synergies in https://www.preprints.org/manuscript/202409.0453/v1
- NFkB related synergies in https://www.preprints.org/manuscript/202409.0696/v1
- TNF related synergies in https://www.preprints.org/manuscript/202409.0471/v1
- DNA repair related synergies in https://www.preprints.org/manuscript/202409.0885/v1
- ABC transporter related synergies in https://www.preprints.org/manuscript/202409.0908/v1
- Interleukin related synergies in https://www.preprints.org/manuscript/202409.1353/v1
- BCL related synergies in https://www.preprints.org/manuscript/202409.0855/v1
- ANTXR2 related synergies in https://www.preprints.org/manuscript/202409.0817/v1
An example of discoveries in 3rd order can be found in -
- WNT10B related synergies in https://www.preprints.org/manuscript/201710.0127/v1
The adaptation of the search engine and the related unpublished research work, which was presented at the Gordon Wnt Research Conference in 2017, in Vermont in USA, is available on Biorxiv at https://www.biorxiv.org/content/10.1101/180927v2 . Adaptation of code for ETC-1922159 data can be found in the link in the above mentioned Biorxiv paper. Please cite the above unpublished manuscript (in Biorxiv) as well as the published manuscript (mentioned at the top) in Oxford University Press, for using this code, if using for research publications. Additionally, those interested in going through the code, can have a pedagogical walkthrough of the same in the unpublished preprint in Preprints at https://www.preprints.org/manuscript/201809.0507/v1
Hope this small contribution will be of help. Thank you for patient hearing.
--
best regards & take care
shriprakash
