Recent introduction of next generation sequencing technology (NGS) has brought revolution to molecular biology and medicine. They also pose new challenges to computational biology and bioinformatics: how to analyze and store data at peta-byte level, how to develop algorithms that fully integrate with the sequencing protocol, and how to interpret the findings. This talk is an overview of our recent work on the bioinformatics of NGS. For DNA-Seq experiments we have developed DRAW (DNA Resequencing Analysis Workflow), a standard analysis pipeline for whole-genome and whole-exome sequencing (WGS/WES) experiments, and SneakPeek, a quality metrics management system for DNA-Seq experiments. DRAW can fully analyze a 350Gbp pair-end whole-exome sequencing dataset in two days with 110 cores on Amazon Elastic Compute Cloud (EC2), and has been fully tested using more than 500 exomes/genomes for human and C. elegans by our lab and our collaborators. DRAW+SneakPeek can be set up on local clusters or cloud environments such as Amazon EC2. We have developed two algorithms for genomic annotation using RNA-Seq data. CoRAL (Classification of RNAs by Analysis of Length) generates biologically interpretable features such as fragment length, cleavage specificity, and antisense transcription from small RNA-seq experiments in order to distinguish between different ncRNA classes. CoRAL could were able to classify six different types of RNA transcripts with ~80% cross-validation accuracy and independent datasets with different tissues. HAMR (High throughput Annotation of Modified Ribonucleotides) performs transcriptome-wide detections of post-transcriptional covalent modifications of ribonucleotides with single-base resolution. HAMR can differentiate between several classes of modifications via a Bayesian model. Using small RNA-seq data we were able to detect 92% of all known human tRNA modification sites that are predicted to affect RT activity, and distinguish between two classes of adenosine and two classes of guanine modifications with 98% and 79% accuracy.