Nb convert param by Congm12 · Pull Request #7 · raphael-group/CalicoST

Congm12 · 2024-07-11T02:37:39Z

Tested on three real data (HT225, HT265, and prostate) and observed a consistent cancer clones and integer copy numbers with the previous convert_params function. All following commands show zero differences.

# prostate cancer sample (first file original, second file with the change)
diff /u/congma/ragr-data/users/congma/Datasets/CalicoST_prostate_example/calicost/clone5_rectangle0_w1.0/cnv_seglevel.tsv /u/congma/ragr-data/users/congma/Datasets/CalicoST_prostate_example/calicost_convertparams/clone5_rectangle0_w1.0/cnv_seglevel.tsv
diff /u/congma/ragr-data/users/congma/Datasets/CalicoST_prostate_example/calicost/clone5_rectangle0_w1.0/clone_labels.tsv /u/congma/ragr-data/users/congma/Datasets/CalicoST_prostate_example/calicost_convertparams/clone5_rectangle0_w1.0/clone_labels.tsv

# HT225 sample (first file original, second file with the change)
diff /u/congma/ragr-data/users/congma/Datasets/WashU_CalicoST/HT225C1_joint/cna_v2/clone4_rectangle0_w1.0/cnv_seglevel.tsv /u/congma/ragr-data/users/congma/Datasets/WashU_CalicoST/HT225C1_joint/cna_v3/clone4_rectangle0_w1.0/cnv_seglevel.tsv
diff /u/congma/ragr-data/users/congma/Datasets/WashU_CalicoST/HT225C1_joint/cna_v2/clone4_rectangle0_w1.0/clone_labels.tsv /u/congma/ragr-data/users/congma/Datasets/WashU_CalicoST/HT225C1_joint/cna_v3/clone4_rectangle0_w1.0/clone_labels.tsv

# HT265 sample (first file original, second file with the change)
diff /u/congma/ragr-data/users/congma/Datasets/WashU_CalicoST/HT265B1-S1H1Fc2U1Z1Bs1/cna_v2/clone4_rectangle0_w1.0/cnv_seglevel.tsv /u/congma/ragr-data/users/congma/Datasets/WashU_CalicoST/HT265B1-S1H1Fc2U1Z1Bs1/cna_v3/clone4_rectangle0_w1.0/cnv_seglevel.tsv
diff /u/congma/ragr-data/users/congma/Datasets/WashU_CalicoST/HT265B1-S1H1Fc2U1Z1Bs1/cna_v2/clone4_rectangle0_w1.0/clone_labels.tsv /u/congma/ragr-data/users/congma/Datasets/WashU_CalicoST/HT265B1-S1H1Fc2U1Z1Bs1/cna_v3/clone4_rectangle0_w1.0/clone_labels.tsv

While comparing the inferred RDR and BAF parameters, cnv_perstate.tsv, we can see differences of magnitude $10^{-5}$.

…c site switch probabilities

…f range

…nces, now the input X should include both observed counts and NB expectation/BB total counts

…allow input argument to initialize log_mu and p_binom

…tering-cnv into main

…runtime of BetaBinom fitting, selecting obs with a high expectation in E step

… step optimization (only using data points with > threshold E step expectation)

…y unique values + number occurrences

…ine_baum_welch

…inding integer copynumber for each HMM state

… probability to enforce reference-based phasing more

…length difference affected by CNV

… utils_IO

…F signals to refine clones

…lmixture_04 version

…length binning by a UMI count threshold). This gives _01 outputs

This reverts commit 5fc50e4.

clean preprocessing and argparse

… parameters

Congm12 and others added 30 commits May 6, 2022 12:16

Create .gitignore

42b428e

Merge branch 'master' into main

86a8365

add HMM with NB and BetaBinom emission probabilities and site-specifi…

e4799e6

…c site switch probabilities

Changed log v to per chromosome

ce49d30

Minor change to backward algorithm

a8c2c63

fix params initialization in Weighted_BetaBinomMixture to avoid out o…

3f86daa

…f range

fix bug in hmm_NB_sharedstates (based on hmmlearn) for multiple seque…

59617a1

…nces, now the input X should include both observed counts and NB expectation/BB total counts

simplify the forward/backward lattice code to use kronecker product; …

40060c0

…allow input argument to initialize log_mu and p_binom

Merge branch 'main' of https://github.com/raphael-group/locality-clus…

b4689e8

…tering-cnv into main

add user defined initialization for NB and BetaBinom mean; to reduce …

bdb8d84

…runtime of BetaBinom fitting, selecting obs with a high expectation in E step

adjust the M step convergence criteria; add a heuristic to speed up M…

583c647

… step optimization (only using data points with > threshold E step expectation)

multi-sample + allow fixing parameters

5ea619b

fix a variable naming bug

32fbfb8

add phase switch utility functions

0304e09

update default parameters in pipeline_baum_welch; shrink input size b…

a0621ed

…y unique values + number occurrences

address the 0 issue in RDR signal; update the initialization in pipel…

ddfc6cf

…ine_baum_welch

add initialize HMM RDR in log/original space option; add gurobi for f…

2167cda

…inding integer copynumber for each HMM state

a trick to speed up M step

141c246

adding a scaling factor of genetic distance in computing phase switch…

db6ef79

… probability to enforce reference-based phasing more

adding rdr-related scale factor for each clone to account for genome …

1087032

…length difference affected by CNV

trying making hmrf an easy-to-run script

caeee7f

revert hmm_NB_BB_phaseswitch temporarily for git pull

6b8aaa5

add back hmm code with scalefactor; fix the directory naming issue in…

f8c748a

… utils_IO

add an example data for running HMM

6e3de6f

gaussian HMM + phase switch for hatchet2

873e2f1

update utility functions

e35c902

a full pipeline for BAF-only clones followed by using both RDR and BA…

b067685

…F signals to refine clones

filter DE genes, better integer copy number inference; leads to norma…

af6053f

…lmixture_04 version

Fixed minor error in initialization_by_gmm method

5ef8887

first phase and bin, and then run HMM without phase switch (variable …

2148326

…length binning by a UMI count threshold). This gives _01 outputs

Congm12 added 29 commits February 19, 2024 18:35

update docs

94db407

add random initialization selection

be9be3b

add simulated tutorial

bc62fcd

add tutorial to readthedocs

5b2e916

move default parsing params to input configurations

4d2ad0c

add example data to github

c7f1e6c

update readme paths for example data

bd5e8c4

fix an edge case

da36c56

simplify environment

d736db0

separate installation into parts

358511d

remove example paths

850fe1c

separate preprocessing from other runs

47e7917

include example parameter configuration file and some reference files

c8ba107

add citation

4d52206

update requirement for output_dir

f4a62fa

source for panels

d938107

updated example data and tutorial

bab0cac

configuration for multiple slices

896c8a0

url for simulation

201d6d2

fix git clone link

d5f1a1e

test

5fc50e4

Revert "test"

e0eefda

This reverts commit 5fc50e4.

update doc

8823977

update example

38f5d4c

update dpc

4c7f960

clean preprocessing and argparse

380983e

Merge pull request #3 from raphael-group/cleanprep

4696325

clean preprocessing and argparse

a more stable version of converting NB mean and dispersion to n and p…

88f6c61

… parameters

switch to the more stable version in distribution fitting fun

e52a69e

Congm12 force-pushed the main branch from f737e8f to c09b026 Compare September 2, 2024 01:34

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Nb convert param#7

Nb convert param#7
Congm12 wants to merge 160 commits intomainfrom
nb_convert_param

Congm12 commented Jul 11, 2024

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

Conversation

Congm12 commented Jul 11, 2024

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants