Fig 1.
Schematic diagram of integrating multilayer perceptron (MLP) with genetics algorithm (GA) for MLP architecture optimization.
Fig 2.
Bayesian tree illustrating the phylogeny of Auricularia species under GTR + G + I model based on the combined sequences of ITS1-5.8S-ITS2 region and RPB2 gene for Auricularia cornea (AJ01) in this study (bold font in the green area) and related species of this genus (in the yellow area), Elmerina efibulata voucher Yuan4525 was considered as outgroup.
The sequences were aligned using MAFFT software and edited using Gblocks program. Bayesian posterior probabilities higher than 0.50 are given for appropriate clades.
Table 1.
Mycelial growth rate (MGR; mm day-1) and days for Auricularia cornea fully colonize the Petri dishes on potato extract agar (PEA), malt extract agar (MEA), yeast extract agar (YEA), hornbeam extract agar (HSA), beech extract agar (BSA) with different carbon sources “fructose, dextrose and maltose” at 25, 28 and 30°C.
Table 2.
Mycelia growth rate (MGR; mm day-1) and spawn run period (SRP) of Auricularia cornea on different substrates obtained from different concentration levels (%) of beech (BS) and hornbeam sawdust (HS), corn flour (CF), wheat (WB) and rice brans (RB) at 25, 28 and 30°C and moisture contents of 65, 75 and 90%.
Fig 3.
Auricularia cornea cultivation in “70% beech sawdust + 30% wheat bran” and “100% wheat bran” with moisture content of 70%, and pH = 6.5.
Table 3.
Effects of different substrate obtained from different concentration levels (%) of beech (BS) and hornbeam sawdust (HS) wheat (WB) and rice brans (RB) on yield, biological efficiency (BE), spawn run period (SRP), days for pinhead formation (DPHF), days for the first harvest (DFFH), number of fruiting body (NFB) and total cultivation period (TCP) of Auricularia cornea.
Table 4.
Statistics on stepwise regression (SR) and multilayer perceptron-genetic algorithm (MLP-GA) models for yield, biological efficiency (BE), spawn run period (SRP), days for pinhead formation (DPHF), days for the first harvest (DFFH), and total cultivation period (TCP) in Auricularia cornea cultivation on different substrates obtained from different ratios of beech and hornbeam sawdust, and wheat and rice brans.
Fig 4.
Prediction of yield, biological efficiency (BE), spawn run period (SRP), days for pinhead formation (DPHF), days for the first harvest (DFFH), and total cultivation period (TCP) in Auricularia cornea cultivation on different substrates obtained from different ratios of beech and hornbeam sawdust supplemented with wheat and rice brans, based on multilayer perceptron-genetic algorithm (MLP-GA) and stepwise regression (SR) models in training subset.
Table 5.
Importance (according to the sensitivity analysis) of the different input variables including different concentration levels of beech (BS) and hornbeam sawdust (HS), and wheat (WB) and rice brans (RB) for achieving maximum yield and biological efficiency (BE), and also minimum spawn run period (SRP), days for pinhead formation (DPHF), days for the first harvest (DFFH), and total cultivation period (TCP) of Auricularia cornea using multilayer perceptron-genetics algorithm models (MLP-GA), and also hidden neuron numbers in each developed model.