Purpose

pysradb allows to retrieve metadata, such as run accession numbers, from SRA and ENA based on multiple criteria:

• Database: SRA or ENA
• Query keywords
• Accession number: a relevant study/experiment/sample/run accession number
• Organism: scientific name of the sample organism
• Library layout: paired or single-end reads
• Sample size: rounded to the nearest megabase
• Publication date
• Sequencing platform: Illumina, Nanopore or PacBio
• Library selection: method used to select and/or enrich the material being sequenced
• Library source: Type of source material that is being sequenced
• Library preparation strategy: sequencing technique intended for the library

Outputs

pysradb generates three different output types:

• Raw metadata file
• Statistics for the search query
• Graphs to illustrate the search results

Sequencing instruments

Comparisons between HiSeq instruments

HiSeq 3000/4000 provides some improvements with respect the previous model HiSeq 2500:

• HiSeq 3000/4000 genere up to 1.5 Tb and 5 Tb reads per run.
• HiSeq 3000/4000 use patterned flow cell technology originally developed for HiSeq X platforms.
• HiSeq 3000/4000 run 3 times faster and yield 65% more reads per lane.
• HiSeq 3000/4000 patterned flow cells contain billions of nanowells at fixed, known positions on the flow cell. The structured organization enables clustering at higher densities compared to non-pattern HiSeq designs.

However, the HiSeq 3000/4000 also have some also some limitations with respect to HiSeq 2500:

• HiSeq 3000/4000 are not recommended for low complexity sequencing. Applications such as non-unique amplicons, 16S, are currently not recommended.
• Libraries with low complexity within the first 25 bases of a read are not expected to produce high quality data.
• Library size restrictions. Libraries that are too long can result in polyclonal clusters that span more than 1 well, these will not pass filter. Smaller libraries will preferentially amplify with Illumina's new kinetic exclusion amplification so tight library distributions ranging from 300-500 bp are recommended.
• Very low tolerance for adapter dimers. Even as little as 1% adapter dimer can take up ~6% of sequencing reads, 10% contamination will take up 84% of reads. Illumina recommends you keep adapter contamination below 0.5% of your entire library.
• Higher duplication rates as compared to HiSeq 2500.
• Low quality read 2 (entire HiSeq 3000 install base is affected).

HiSeq 3000/4000 support DNA-seq, RNA-seq , ChIP-Seq, mate-pair, small RNA and exome library preparation. Any library preparation where there is enough sequence diversity is currently supported. Amplicon, 16S and applications with low sequencing diversity are currently not supported on the HiSeq 3000 / 4000.

HiSeq 2500 is considered the most reliable model according to different sources.

What type of read quality is expected from the HiSeq 3000/4000 ?

• 2 x 50bp ≥85% bases > Q30
• 2 x 75bp ≥80% bases > Q30
• 2 x 150bp ≥75% of bases >Q30

What is the difference between MiSeq and HiSeq?

HiSeq and MiSeq platforms are among the most widely used platform to study microbial communities. But the two platforms differ in the length and amount of reads. MiSeq can run 600 cycles to produce 200 million 300 bp reads, on the other hand, HiSeq 2500 can run 500 cycles to produce 120 million 250 bp.

What are the differences between HiSeq and NovaSeq?

The Illumina NovaSeq provides a massive upgrade in sequencing throughput compared to the HiSeq 4000. There are more stringent library requirements and requires a larger sample size. Due to the vast amount of data produced by the NovaSeq and the known issue of index swapping, unique dual-indexed libraries are required.

What are the characteristics of HiSeq X instruments?

• HiSeq X is recommended for whole genome sequencing only (including whole bisulfite sequencing). This means that it is not adequate for RNA-seq, exome, ChIP-seq or small RNA-seq applications.
• Plant and animal samples can be sequenced on the HiSeq X.
• Expect coverate is over 30x or approximately 375 million reads per lane by loading one sample per lane.
• Hiseq X Ten generates utilize 2x150 base pair read configurations and has slightly better GC coverage than the HiSeq 2500.

What are the differences between MiSeq and Nextseq?

The NextSeq Series of systems delivers the power of high-throughput sequencing with the simplicity of a desktop sequencer. NextSeq instruments represent an improvement when compared with Miseq, despite generating sorter reads (150bp, compared to MiSeq 250bp). NextSeq is recommended in the following applications & methods:

• Exome & large panel sequencing (enrichment-based)
• Single-cell profiling (scRNA-Seq, scDNA-Seq, oligo tagging assays)
• Transcriptome sequencing (total RNA-Seq, mRNA-Seq, gene expression profiling)
• Methylation sequencing
• Metagenomic profiling (shotgun metagenomics, metatranscriptomics)
• Cell-free sequencing & liquid biopsy analysis

Regarding the maximum number of reads per ran, MiSeq can generate 25 million, vs 400 million generated by the Nextseq 550 instrument. MiSeq recommended for sequencing samples of low diversity.

What are the differences between HiSeq and NextSeq?

The main technical difference between HiSeq and NextSeq will be the number of dyes each machines use. HiSeq uses traditional color coding with four different dyes, while NextSeq uses two dyes. This does not give any practical differences in terms of the data quality, but the trend in illumina sequencers are more into the direction of reducing the number of dyes.

What is the difference between Nextseq and NovaSeq?

The NovaSeq 6000 system offers deep and broad coverage and is recommended for large whole-genome sequencing (human, plant, animal) projects. It generates 250 bp reads, with 20 billion maximum reads per run. NovaSeq 6000 instruments have not application based restrictions.

Illumina maximum read-length summary

• MiSeq: between 300 and 600 bp
• NextSeq: 300 bp
• HiSeq 2500: between 250 and 500 bp (depending of the sofware)
• HiSeq 4000: 150 bp
• HiSeq X: 150 bp

Nanopore models - single-molecule ultra-long-read sequencing

Nanopore sequencing provides the longest read lengths, from 500 bp to the current record of 2.3 Mb, with 10-30-kb genomic libraries being common. Even after error correction, sequencing error rates of corrected nanopore reads (1.5-9%) are still higher than those of corrected PacBio reads (<1%).

PacBio SMRT instruments - single-molecule long-read low-error rate sequencing

PacBio Sequel II CLR sequencing represents a major advancement in sequencing throughput over previous PacBio platforms with the production of more sequencing data and longer reads versus RS II and the Sequel I. The PacBio HiFi sequencing method yields highly accurate long-read sequencing datasets with read lengths averaging 10-25 kb and accuracies greater than 99.5%.