Active machine learning-driven experimentation to determine compound effects on protein patterns
Abstract
High throughput screening determines the effects of many conditions on a given biological target. Currently, to estimate the effects of those conditions on other targets requires either strong modeling assumptions (e.g. similarities among targets) or separate screens. Ideally, data-driven experimentation could be used to learn accurate models for many conditions and targets without doing all possible experiments. We have previously described an active machine learning algorithm that can iteratively choose small sets of experiments to learn models of multiple effects. We now show that, with no prior knowledge and with liquid handling robotics and automated microscopy under its control, this learner accurately learned the effects of 48 chemical compounds on the subcellular localization of 48 proteins while performing only 29% of all possible experiments. The results represent the first practical demonstration of the utility of active learning-driven biological experimentation in which the set of possible phenotypes is unknown in advance.
Article and author information
Author details
Copyright
© 2016, Naik et al.
This article is distributed under the terms of the Creative Commons Attribution License permitting unrestricted use and redistribution provided that the original author and source are credited.
Metrics
-
- 7,071
- views
-
- 1,230
- downloads
-
- 42
- citations
Views, downloads and citations are aggregated across all versions of this paper published by eLife.
Download links
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
Further reading
-
- Cell Biology
The primary cilium is a microtubule-based organelle that cycles through assembly and disassembly. In many cell types, formation of the cilium is initiated by recruitment of ciliary vesicles to the distal appendage of the mother centriole. However, the distal appendage mechanism that directly captures ciliary vesicles is yet to be identified. In an accompanying paper, we show that the distal appendage protein, CEP89, is important for the ciliary vesicle recruitment, but not for other steps of cilium formation (Tomoharu Kanie, Love, Fisher, Gustavsson, & Jackson, 2023). The lack of a membrane binding motif in CEP89 suggests that it may indirectly recruit ciliary vesicles via another binding partner. Here, we identify Neuronal Calcium Sensor-1 (NCS1) as a stoichiometric interactor of CEP89. NCS1 localizes to the position between CEP89 and a ciliary vesicle marker, RAB34, at the distal appendage. This localization was completely abolished in CEP89 knockouts, suggesting that CEP89 recruits NCS1 to the distal appendage. Similarly to CEP89 knockouts, ciliary vesicle recruitment as well as subsequent cilium formation was perturbed in NCS1 knockout cells. The ability of NCS1 to recruit the ciliary vesicle is dependent on its myristoylation motif and NCS1 knockout cells expressing a myristoylation defective mutant failed to rescue the vesicle recruitment defect despite localizing properly to the centriole. In sum, our analysis reveals the first known mechanism for how the distal appendage recruits the ciliary vesicles.
-
- Biochemistry and Chemical Biology
- Cell Biology
Stem cell differentiation involves a global increase in protein synthesis to meet the demands of specialized cell types. However, the molecular mechanisms underlying this translational burst and the involvement of initiation factors remains largely unknown. Here, we investigate the role of eukaryotic initiation factor 3 (eIF3) in early differentiation of human pluripotent stem cell (hPSC)-derived neural progenitor cells (NPCs). Using Quick-irCLIP and alternative polyadenylation (APA) Seq, we show eIF3 crosslinks predominantly with 3’ untranslated region (3’-UTR) termini of multiple mRNA isoforms, adjacent to the poly(A) tail. Furthermore, we find that eIF3 engagement at 3’-UTR ends is dependent on polyadenylation. High eIF3 crosslinking at 3’-UTR termini of mRNAs correlates with high translational activity, as determined by ribosome profiling, but not with translational efficiency. The results presented here show that eIF3 engages with 3’-UTR termini of highly translated mRNAs, likely reflecting a general rather than specific regulatory function of eIF3, and supporting a role of mRNA circularization in the mechanisms governing mRNA translation.