ChromImpute(1)

✨ NAME ✨

okay bestie, meet ChromImpute 🧚 It's the General Commands Manual's ChromImpute(1) page, and it's all about large-scale systematic epigenome imputation 💅

🌸 SYNOPSIS 🌸

Here's how you summon it 🪄

java -Xmx[GB]g -jar /usr/share/java/chromimpute.jar [options]

💜 DESCRIPTION 💜

okay so ChromImpute is software for large-scale systematic epigenome imputation, and it's literally so iconic 😭 It takes an existing compendium of epigenomic data and uses it to predict signal tracks for mark-sample combinations that were never experimentally mapped. It can also make a potentially more robust version of data sets that were mapped experimentally 🔥

The predictions are based on features from signal tracks of other marks mapped in the target sample, plus the target mark in other samples. Those features get combined using an ensemble of regression trees, and the whole thing is giving main character energy 🌈

🦋 OPTIONS 🦋

• Convert — turns signal tracks into binned signal resolution tracks 🍓
• ComputeGlobalDist — computes the global distance based on correlation for each mark in each sample with the same mark in all other samples. It creates a file for each mark in each sample with a ranked list of the globally nearest samples 🧠
• GenerateTrainData — takes a directory of converted data and global distances, then generates a set of training data instances 📚
• Train — trains regression trees based on the feature data produced by GenerateTrainData 🌳
• Apply — applies the predictors generated in the Train command to generate the imputed data ✨
• Eval — compares the agreement between an observed and imputed data set 🫶
• Version — gives you the version of ChromImpute 🎀

🔗 SEE ALSO 🔗

Go check out the vibes here 👀

• http://www.biolchem.ucla.edu/labs/ernst/ChromImpute/

👑 AUTHOR 👑

ChromImpute was written by Jason Ernst 💖 The queen behavior is real 💅

That's the tea, bestie 🫖 ChromImpute 1.0.3, December 2017 🦋🌸 stay iconic 💜✨