diff --git a/README.md b/README.md index 0985d4f..a554254 100644 --- a/README.md +++ b/README.md @@ -1,6 +1,6 @@ # SIMPA - Single-cell chIp-seq iMPutAtion -SIMPA is a method for **S**ingle-cell Ch**I**P-seq i**MP**ut**A**tion that leverages predictive information within epigenomic data from ENCODE to impute missing protein-DNA interactions for a histone mark or transcription factor of interest. SIMPA was tested on a recent dataset (Grosselin et al. 2019) to impute missing regions in sparse data from single-cell ChIP-seq of H3K4me3 and H3K27me3 in B-cells and T-cells. Different to common single-cell imputation methods, SIMPA leverages predictive information within bulk ChIP-seq experimental data. This dataset contains > 2.200 experiments downloaded from ENCODE and available in this repository, preprocessed for SIMPA. The user provides peaks of one single cell that are used by SIMPA to impute missing interactions for the target (histone mark or transcription factor, also specified by the user) of interest while keeping cell-type specificity and the cells individuality. In our preprint on [bioRxiv](https://www.biorxiv.org/content/10.1101/2019.12.20.883983v3) we present SIMPA's capability to complete sparse single-cell input while improving cell-type clustering and recovering cell-type-specific pathways. Moreover, SIMPA was extended by InterSIMPA which allow you to interpret the underlying machine learning models trained for the purpose of imputation. An example for how to do this, is provided below. +SIMPA is a method for **S**ingle-cell Ch**I**P-seq i**MP**ut**A**tion that leverages predictive information within epigenomic data from ENCODE to impute missing protein-DNA interactions for a histone mark or transcription factor of interest. SIMPA was tested on a recent dataset (Grosselin et al. 2019) to impute missing regions in sparse data from single-cell ChIP-seq of H3K4me3 and H3K27me3 in B-cells and T-cells. Different to common single-cell imputation methods, SIMPA leverages predictive information within bulk ChIP-seq experimental data. This dataset contains > 2.200 experiments downloaded from ENCODE and available in this repository, preprocessed for SIMPA. The user provides peaks of one single cell that are used by SIMPA to impute missing interactions for the target (histone mark or transcription factor, also specified by the user) of interest while keeping cell-type specificity and the cells individuality. In our preprint on [bioRxiv](https://www.biorxiv.org/content/10.1101/2019.12.20.883983v3) we present SIMPA's capability to complete sparse single-cell input while maintaining cell-type clustering and recovering cell-type-specific pathways. Moreover, SIMPA was extended by InterSIMPA which allow you to interpret the underlying machine learning models trained for the purpose of imputation. An example for how to do this, is provided below.