mcavallaro / gloop

Software supporting M. Cavallaro, et al., 3′-5′ crosstalk contributes to transcriptional bursting. Genome Biol 22, 56 (2021). https://doi.org/10.1186/s13059-020-02227-5

Gloop

This repository contains supporting software to reference [1]. Please cite [1] if you find this repository useful. The software is organised as follows.

• R scripts for tidying flow cytometry .fcs data and resolving cell-cycle stages (G1/M/G2).

  • clustering_2.R
  • clustering_3.R
  • clustering_caller.R
  • compensate_caller.R
  • CompensateFlowSet.R

  These require flowCore and flowClust [2,3].

• c++ implementation of the Gillespie algorithm for the simulation of gene expression based of the reaction network. It requires the GNU Scientific Library (GSL) ver. 2> [4]. By default it saves the simulation results into a directory named .\results). A makefile is provided for initial setting, compilation, and linking with gcc.

  make install
  make
  

  Simulation parameters are passed from STDIN, e.g.:

   ./main.exe t N $\alpha$ $\beta$ d $\lambda_{on}$ $\lambda_{on}$ l
  

• MCMC samplers implemented in python and pymc [5] for the three phenomelogical models described in [1]:

  • BetaPoissonModel.py
  • NegativeBinomialModel.py
  • PoissonModel.py

  these can be launched as:

  python BetaPoissonModel.py data_file_name $\mu_X$ $se_{\mu_X}$ BCK_params_file_name
  

  BCK_params_file_name contains the parameters obtained from the controll cell.

  Some diagonistic methods are imported from pymc3 [6]. The sampler for the model with no measurement equation and the utils to rescale and tidy the calibration data are in separated files:

  • NegativeBinomialModel_no_error.py
  • utils.py

• R and bash scripts for the bioinformatic interrogation of ChIA-PET data to extract 3'-5' interaction scores genome wide:

  • Download_trim_chia_pet2.sh
  • makehicmatrix.sh
  • calculate_3_5_interaction_res_2000.r
  • loopscore_functions.r

[1] M. Cavallaro, M.D. Walsh, M. Jones, et al., 3'-5' interactions contribute to transcriptional bursting. Genome Biol 22, 56 (2021). https://doi.org/10.1186/s13059-020-02227-5

[2] F. Hahne, et al., flowCore: a Bioconductor package for high throughput flow cytometry. BMC Bioinformatics 10 106 (2009). http://www.ncbi.nlm.nih.gov/pubmed/19358741

[3] K. Lo et al., flowClust: a Bioconductor package for automated gating of flow cytometry data. BMC Bioinformatics 10, 145 (2009). https://www.ncbi.nlm.nih.gov/pubmed/19442304

[4] M. Galassi et al., GNU Scientific Library Reference Manual (3rd Ed.), ISBN 0954612078. http://www.gnu.org/software/gsl/

[5] A. Patil, D. Huard, C. Fonnesbeck, PyMC: Bayesian Stochastic Modelling in Python. J. Stat. Softw. 35 1–81 (2010). https://pymc-devs.github.io/pymc/

[6] J. Salvatier, T. Wiecki, C. Fonnesbeck, Probabilistic Programming in Python using PyMC3. PeerJ Comput. Sci. 2, e55 (2015). https://docs.pymc.io/