Machine learning based de novo assembly, alignment and annotation

Are there any ML (machine learning) software tools for de novo assembly, alignment and variant calling/annotation on any Galaxy instance? Have read that they perform better metrics than conventional.