Back

Comprehensive Sequencing of Environmental RNA from Japanese Medaka at Various Size Fractions and Comparison with Skin RNA

Hiki, K.; Jo, T. S.

2024-09-25 molecular biology
10.1101/2024.09.23.614187 bioRxiv
Show abstract

Environmental RNA (eRNA) is emerging as a non-invasive tool for assessing the health of macro-organisms, but key information on its origin and particle sizes remains unclear. In this study, we performed comprehensive RNA-sequencing of eRNA (> 13 Gb/sample) collected from tank water containing Japanese medaka (Oryzias latipes), using sequential filtration through filters with pore sizes of 10, 3, and 0.4 m. Fish skin RNA was also sequenced to reveal the origin of eRNA. Our results showed that the 3-10 m fraction contained the lowest relative abundance of microbial RNA, the highest amount of medaka eRNA, and the largest number of detected medaka genes (5398 genes), while the 0.4-3 m fraction had the fewest (972 genes). Only a small number of genes (42 genes) were unique to the 0.4-3 m fraction. These findings suggest that a 3 m filter is optimal for eRNA analysis, as it allows for larger filtration volumes while maintaining the relative abundance of macro-organism eRNA. Furthermore, 81% of the genes detected in eRNA overlapped with skin RNA, indicating skin is a major source of fish eRNA. SynopsisThe 3 {micro}m filter is recommended to maximize the detection of eRNA from macro-organisms while reducing the relative abundance of microbial RNA. TOC Art O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=77 SRC="FIGDIR/small/614187v2_ufig1.gif" ALT="Figure 1"> View larger version (21K): org.highwire.dtl.DTLVardef@1702cc9org.highwire.dtl.DTLVardef@18c4d2borg.highwire.dtl.DTLVardef@1e4c13org.highwire.dtl.DTLVardef@114a4b1_HPS_FORMAT_FIGEXP M_FIG C_FIG

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.